Every forum answer to this is somebody’s opinion. It is not an opinion. It is arithmetic with a priority order, and once you know the order you can usually put the offset back to nothing without moving a single pipe.
Search this and the whole first page of results is forum threads. Somebody says it drives them mad, somebody else says nobody notices, a third person says theirs has been four inches out for eleven years and they have never once looked at it. All of that is true and none of it helps, because the question is not whether asymmetry bothers you. It is how many inches you are out, how many of those inches the run can absorb before anything visible changes, and which of four moves you spend them on. That is a calculation, it takes about ten minutes, and it has a correct order.
A cabinet shop settles this in one line, and it is worth memorising exactly as written: the faucet centres on the window, the bowl centres on the faucet, and the box is whatever it has to be. Three objects, three rules, in that order. Almost every argument about off-centre sinks is two people applying the three rules in different orders and not realising it.
What they search at midnight “sink not centered under window” — nearly always the night before the countertop is templated, when the cabinets are already screwed to the wall and the number that is wrong is finally visible.
The faucet is the reference because it is the only one of the three objects that shares air with the window. A counter is 36″ off the floor and a sill above a run typically lands somewhere between 40″ and 48″. The bowl lives below the counter line; the cabinet lives below that. Neither of them is ever in the same field of view as the glass. The faucet, on the other hand, is a vertical object roughly 12″ to 18″ tall standing directly in front of the opening, silhouetted against the brightest thing in the room. It is at eye level. It is the thing your eye lands on when you walk in the door, and it is the thing a photograph of that wall is actually about.
That silhouette is also why the error is so easy to see. The eye compares two gaps that sit side by side far more unforgivingly than it judges any single object against a centre line it cannot see. Put a faucet between two casing edges and you have made exactly that comparison unavoidable — and, as the next section shows, the comparison doubles the error before you ever get to it.
The bowl centres on the faucet because the faucet has to deliver water into it. On a single bowl, the deck hole and the bowl centre are the same line, so fixing one fixes the other. That is not a constraint so much as a definition.
The box is whatever it has to be because nothing about a base cabinet is visible at the resolution the window comparison happens at. The joint between two base cabinets is a reveal of an eighth of an inch at knee height, below the counter, in shade, usually with a door handle near it. Nobody has ever walked into a kitchen and noticed that the sink base was two inches off the bowl. The only hard limit is that the bowl has to physically fit inside the box, and that limit is tighter than people expect: a frameless base loses ¾″ per side to the panels, so a 33″ sink in a 36″ base sits with about ¾″ of air each side — and that air is the fabricator’s tolerance and the installer’s hands, not travel. Sizing the base to the bowl is its own subject — take that as settled before you start this page. What it means here is that moving the bowl inside its own base is not one of the moves you have.
Which surfaces a quiet fifth option that belongs here rather than in the list of four: a bigger sink base buys the bowl room to move. Sink bases come at 24, 30, 33, 36 and 42″. Put that same 33″ sink in a 42″ base and the interior is about 40½″, so even after leaving the same ¾″ of working air at each panel the bowl can sit anywhere across about 3″ either side of the box centre. One cabinet size up hands you most of the correction range on this page without touching anything else. What you give up is the drawer or the door those inches would otherwise have been, and it is the single most underused fix in the list.
Here is the piece of arithmetic that reframes the whole problem. You are not looking at the offset. You are looking at the difference between two gaps, and that difference is twice the offset. Move the faucet 1″ to the right of the window centre and the gap on its left grows by an inch while the gap on its right shrinks by an inch. What the eye compares is 2″. Move it 2″ and the comparison is 4″.
With that doubling in hand, the thresholds stop being a matter of taste:
| Offset | Gap difference seen | Against a bare drywall return | Against a trimmed window | Watch for |
|---|---|---|---|---|
| ¼″ | ½″ | Invisible | Invisible | Below the accuracy of the install anyway |
| ½″ | 1″ | Invisible | Findable if you look for it | This is the stop-worrying line. Do not spend a cabinet on it. |
| 1″ | 2″ | Noticed by some people, some days | Read as wrong by most people once pointed out | The zone where people feel it without being able to name it |
| 2″ | 4″ | Clearly off | Obviously a mistake | Fix it now. This is the version that gets photographed and posted. |
| 4″ and up | 8″+ | Reads as a failed attempt at symmetry | Same, louder | Either correct it fully or go far enough that it reads as a decision |
Why the trim matters so much: a casing gives the eye two hard, high-contrast vertical lines a known distance apart, and the faucet stem becomes a third vertical between them. Three parallel lines, two of which are definitionally symmetric, is the exact stimulus the visual system is best at auditing. Take the casing away and a drywall return does the opposite — the corner is a rounded bead, light falls off across it as a soft gradient, and there is no crisp edge to measure from. The same half inch simply dissolves. So the answer to “how much can I get away with” depends on a trim decision that is usually made months later by somebody else. If the window is going to be cased, hold yourself to ½″. If the returns are drywall and staying that way, 1″ is genuinely fine.
Two things also make an offset less visible, and both are worth knowing before you set about fixing it. Viewing angle is the first: asymmetry is at its worst seen square-on from the opposite side of the room, and it softens fast once you are looking at the wall obliquely. Stand in the doorway you actually enter by before you decide. The second is competing verticals — a tall pantry, a fridge, a pair of pendants, a range hood. A wall with six strong verticals on it does not invite the two-gap comparison the way an empty wall with one window does.
And the honest last word on the tolerance: there is an ugly middle. Under about ½″ nobody sees it. Beyond about half the width of the window, the faucet has visibly left the opening and reads as sitting beside it rather than failing to be centred in it. It is everything between those two — the near misses — that looks like an error, because it looks like somebody tried.
You have some number of inches to lose and four places to put them. Work down the list in order, because the list is ordered by how much has to change and by who has to be told.
First to try, and the only one where nothing is ordered differently
A run that dies into a wall at each end has a filler at each end, and it must — walls are not straight, corners are not square, and a run built to the exact wall measurement does not go in. Those fillers are scribe material, and they are also your entire correction range. Base fillers come as a 3″ board and a 6″ board (BF330 and BF630), and both get ripped down on site. Make the left one narrower and the right one wider by the same amount, and the entire run slides sideways by that amount with nothing else changing.
The trap is the ceiling, not the floor. The two fillers have to add up to the same total, so your travel is limited by how far the growing one can grow — and a base filler stops at 6″. Design a wall-to-wall run with 6″ of filler at each end and you have exactly zero travel, because neither can get any wider. Design it with 2″ at one end and 6″ at the other and you can move 2″ in one direction and nothing in the other. The practical floor is about ½″: below that a filler is too thin to scribe to a wavy wall and too fragile to rip cleanly.
The exact fix, and the one that changes an order line
Cabinet widths step in 3″ from 12″ to 36″ and then jump straight to 42″ — there is no 39″ box — which is why the offsets people end up with are so often 1″, 2″ or 4″: the grid cannot land anywhere else. Retyping a width breaks the grid on purpose. Shrink one box on the window side of the sink by the amount you are out and the sink base, and everything after it, walks along the wall by exactly that much.
Pick the box carefully. Retype a cabinet that has no twin nearby, because what actually shows is a door 2″ narrower than the identical door beside it. A box at the far end of the run, a single door base tucked against a corner, or a box whose neighbour is a dishwasher or a fridge bay are all invisible. The worst choice is one of a matched pair flanking the range, which is the one place on the wall where somebody is comparing two gaps. Do not retype the sink base itself — you already sized that to the bowl — and do not retype anything that has a fixed appliance opening inside it.
In our planner you can type any width you like onto a cabinet, and the quote then marks that line as resized rather than treating it as a catalogue item. That is the honest signal: a retyped box is a made-to-size box, and it gets handled as one.
Free, when one end of the run is open
Leave every box at its catalogue width, move the lot sideways, and let the far end come up short. If that end is open — a doorway, a fridge bay, a return run, a walkway — the gap is not a gap, it is just the run ending a little earlier. A fridge bay that grows from 36″ to 40″ looks exactly like a fridge bay. A run that stops 4″ sooner beside a doorway looks like a run that stops beside a doorway.
If the far end dies into a wall, this move collapses back into Move 01 and inherits the 6″ filler ceiling, because the gap has to be filled by something. Finish an open end with a finished end panel rather than leaving a raw side on show, and remember the countertop does not have to stop where the cabinets do — a small overhang past the last box reads as intentional.
Last, and usually unnecessary — see how little it actually has to move
This is the move everyone on the forums reaches for first and the one that matters least, because for the corrections on this page the plumbing very often does not have to move at all. A sink base is usually 30″ to 42″ wide; the stub-out only has to be somewhere behind it. Slide the box 4″ and the existing stub is still comfortably inside the cabinet. The trap arm takes a 4″ sideways jog and gives up under a tenth of an inch of fall doing it.
What moving the plumbing genuinely gets you is the freedom to put the sink anywhere on that wall — a far larger job than any correction on this page needs. Section 06 lays out what that involves, step by step. Reach for it when the three moves above have run out, not before.
| Move | What physically changes | Range it covers | Effort | The catch |
|---|---|---|---|---|
| Split the fillers | Two saw cuts | Up to 6″, if the travel was designed in | Minutes, on site | Dead on arrival if both fillers are already at maximum |
| Retype a box | One cabinet ordered at a typed width | Any amount, exactly | A typed number, before the order | The line is flagged as resized, and the inches reappear elsewhere |
| Shift the run | Nothing — everything just sits further along | Any amount, if one end is open | None | Only exists when an end is open; otherwise it is Move 01 |
| Move the plumbing | A stub-out, and possibly a wall | Anything, including the whole wall | Trivial while the wall is open; the most disruptive of the four once it is closed | Vents and drain falls are licensed work — confirm locally |
The order is not arbitrary. Moves 01 and 03 change nothing that anybody will ever see. Move 02 changes one line on an order. Move 04 changes the building. Start at the top of that list and stop as soon as the number is small enough.
A surprising share of off-centre sinks were centred correctly on paper. The drawing was right and the wall was different, because two numbers on it came from two different datums.
Half-inch drywall on a return takes ½″ off each end of a wall, so a room that framed at 145″ finishes at 144″. If you take the wall length off the framing plan and the window position off a tape held against finished drywall, you are half an inch out before you start, and the two errors do not cancel — they add or subtract depending on which end you measured from. One tape, one corner, one direction, everything recorded as a distance from that same corner. Not “the window is 40″ wide and roughly in the middle”, but “left casing edge at 48″, right casing edge at 88″, from the left-hand finished corner.” A running dimension chain from one datum cannot drift; a string of individual widths always does. The three-heights method for taking the wall itself is the companion to this and belongs in the same session.
Then there is the second question, which almost nobody asks and which decides the answer: which centre are you centring on? A window has at least two, and they are frequently not the same number.
They diverge for ordinary reasons. A window near a corner often has its casing ripped narrower on the corner side because there was not room for a full width. A slider has a fixed half and a sliding half that are not the same width. A casement’s hinge stile and lock stile differ. A window with a mulled sidelight has a centre of glass nowhere near its centre of trim.
The working rule: centre on the strongest pair of vertical lines in the assembly. On a trimmed window that is the outside edges of the casing. On a drywall-returned window it is the two return corners. On a big picture window in a thin frame it is the glass. If the strongest line in the window is a single vertical — a meeting stile, a mullion — then you are no longer centring between two things, you are aligning to one, and you should decide deliberately whether the faucet lands on that line or clearly away from it. Landing an inch off it is the worst of both.
One more measurement worth thirty seconds: take the casing edges at the head and again down at the sill. If the two centres differ, the window is not plumb in the wall, which means your target is a range rather than a point. That is good news. Centre on the average, stop optimising, and go and spend the effort somewhere it is measurable.
A double bowl is usually chosen for how it works. It is also the most useful geometric tool on this page, because it gives you a second centre to aim with.
On a single bowl, the faucet, the bowl centre and the visual centre are all the same line. There is nothing to adjust. On a double, those separate: there is the deck centre, the divider line, and the centre of each bowl. Which of them you put under the window is now a decision, and an offset double — 60/40, 70/30, whatever the maker calls it — spreads them further apart and gives you more to work with.
Here is the arithmetic that makes it matter, and it is at the small end, not the large. Take a 33″ sink. Deduct the two outer walls and the divider and you have roughly 30″ of bowl to divide. As an equal double that is two bowls of just under 15″, with the divider dead on the sink centre line. Now shift the faucet 4″ to one side to catch the window: the stream lands about 3½″ off the divider wall, in the inner quarter of a 15″ bowl rather than anywhere near its middle. A hard tap throws water up the divider and back out onto the deck, every time.
Take the same 33″ sink as a 60/40 and the large bowl is about 18″ wide, straddling the sink centre, with its own centre roughly 6″ toward the wide side. That same 4″ faucet shift now lands comfortably inside the large bowl, well away from any wall. Nothing moved except which sink you ordered.
Using it deliberately means three choices, in this order:
Two limits to be straight about. A low-divide sink, where the divider sits a few inches below the rim, does not help here — the divider is still in the same plane, still lines up or does not. And an offset double is a real working compromise in its own right: you get a bowl that takes a sheet pan and a second bowl that mostly holds a colander. Use it because it suits how you wash and take the geometry as a bonus. Choosing the sink you do not want in order to gain 3″ of faucet travel is a bad trade, when Move 01 is a saw cut.
“Move the plumbing” covers five very different jobs, and lumping them together is why the forum answers are so frightening. Here they are in order of escalation. Most kitchens never get past the second.
The rough thresholds, which are the useful takeaway: under about 3″, nothing behind the wall changes at all — the trap and the connectors absorb it without anybody thinking about it. Out to roughly a foot, a longer trap arm generally does it, assuming the fall and the arm length work. Past that, you are into stub-out territory, and the question becomes whether opening one bay of wall is easier than retyping one cabinet. While the wall is open, it usually is. Once it is closed and tiled, it never is.
Everything in this section is planning arithmetic, not a code answer. Drain falls, trap-arm lengths, venting and whether any of it needs a permit vary by province and municipality, and they are the part of a kitchen that quietly causes damage when they are wrong. Confirm with a licensed plumber and your local inspector before anything is cut. What you can do without one is exactly what the first three sections of this page describe — move the cabinets instead.
One sequencing note that saves more grief than any of the above. If the wall is open right now, take fifteen minutes and mark the sink centre on the framing in pencil, at the position you have actually calculated. The single most common version of this problem is a stub-out placed where a plumber assumed the sink would go, months before anybody drew a cabinet run — usually centred on the window, which sounds right and is only right if the run happens to co-operate.
The priority order in Section 01 assumes one window, roughly central, with wall on both sides. Four common situations break that, and each has a different right answer.
A window that is off-centre in the wall. This is the case that makes people abandon the whole exercise, and it should not. Centring the sink on the window necessarily puts the sink off-centre in the run, and that is fine, because nobody centres a sink on a run — there is no line to compare it to. What the offset does threaten is landing space. A sink wants clear counter on both sides, generously on the side dirty dishes arrive; if chasing the window leaves under about 18″ on one side, the window loses. Landing space is used every day and symmetry is looked at. Centre on the run instead, and then go far enough that it reads as a decision rather than a near miss.
Two windows. Centre on the pair, using the outermost casing edges as your two lines — that is what the eye brackets. Do not centre on the pier between them unless the pier is genuinely the dominant element, because a faucet dead in front of a strip of wall is the one position where the silhouette disappears entirely. Some people like that and prefer the view unobstructed. Most find it removes the point of putting the sink under a window at all. If the two windows are unequal, use the outer edges rather than the pier centre — visual mass, not construction geometry.
Corner windows. Two separate problems arrive together. First, there may be no wall at the end of the run for the counter to die into, and a countertop cannot be scribed to a window frame the way it can to drywall — a frame is not flat, is not plumb where two units mull together, and moves seasonally. Always leave a deliberate reveal or a filler there; a tight fit to a window frame is a fit that opens. Second, a corner window has no meaningful centre. Pick the pane the sink actually faces and centre on that, or centre on the run and let the window be a window. Nobody reads symmetry into glass that wraps a corner.
A window over a peninsula or an island. There is no wall to measure from and no casing to compare against, so the run is the only reference there is. Centre the sink on the island. The layout mistakes page covers what else that decision drags with it.
The thread running through all four: when the window stops giving you a usable centre line, stop treating it as the reference and promote the next strongest line on the wall — the run, the cabinet group, the range opposite. What you must not do is keep aiming at a centre that is not there and land two inches off it.
A 12′ wall — 144″ of finished drywall, wall at both ends. The run, left to right, as drawn:
| Position | Item | Width | Runs from | To |
|---|---|---|---|---|
| 1 | Base filler BF630, cut | 6″ | 0″ | 6″ |
| 2 | B18 base | 18″ | 6″ | 24″ |
| 3 | DB30-3 drawer base | 30″ | 24″ | 54″ |
| 4 | SB36 sink base | 36″ | 54″ | 90″ |
| 5 | Dishwasher opening | 24″ | 90″ | 114″ |
| 6 | B27 base | 27″ | 114″ | 141″ |
| 7 | Base filler BF330, cut | 3″ | 141″ | 144″ |
The sink base spans 54″ to 90″, so its centre — and the bowl’s, and the faucet’s — is at 72″, which is also the centre of the wall. Tidy. Then you measure the window from the same left-hand corner: outside casing edges at 48″ and 88″, so a 40″ casing centred on 68″.
You are 4″ out, and the gaps either side of the faucet will differ by 8″. Four resolutions:
The lesson underneath the numbers is resolution one’s ceiling. That run had 9″ of filler in it and could still only correct 3″, because correction range is limited by the filler that has to grow, and base fillers stop at 6″. You fix your correction range at the moment you decide where the fillers go, which is well before you know you need one. Where a window is involved, put the bigger filler at the end the sink is likely to travel away from, and keep the other end lean.
All of this is decidable on paper, and all of it is much easier to get wrong on paper than you would think, because a plan view flattens the one thing you are judging. Looking down at a run from above, the window is a gap in a line; the faucet is a dot. The comparison your eye will eventually make does not exist in that view.
So look at the wall straight on. In our planner that is the elevation view, which draws the run against the wall — cabinets, sink, window, all in the same picture, from the position you will actually be standing in. That is where a 2″ miss becomes obvious and a 4″ miss becomes embarrassing, months before anything is built.
Three things do the work here. The elevation view draws the run against the wall, so you see the sink against the window instead of imagining it. The Measure tool reads the distance between two points you click, and it works on the elevation canvas as well as the plan — click the window’s left edge to the sink base’s left side, then the window’s right edge to its right side, and compare the two numbers. Equal numbers mean the box is centred on the opening; the difference between them is the doubled error your eye will meet later. The planner draws the window rather than its casing, and the tap only exists in the 3D view, so measure to the box here and eyeball the tap in 3D. And the GAP number between cabinets can be clicked and retyped, which is how you set a filler to 3″ or 5½″ exactly rather than sketching it. Typing a non-catalogue width onto a cabinet is allowed too, and the Quote page then marks that line as resized. How to drive all three is here.
Nothing in the planner validates the layout. There is no clearance readout, nothing checks the sink against the window, and no warning appears if your faucet is four inches off. The Measure tool reads between two points you click, and the GAP number is a number you type — both are manual, both answer the question only if you ask it. The software will draw a badly centred sink perfectly happily. The check in the callout above is a thing you do.
The sequence that actually works, and the order matters:
If you do that once, properly, you will either fix the offset entirely or you will find out that you cannot — and knowing which, before the cabinets arrive, is the whole of the win. The people on those forum threads are not wrong about how they feel. They are just all answering three months too late.
No, but the faucet should be, and it is usually achievable. The shop rule is that the faucet centres on the window, the bowl centres on the faucet, and the cabinet is whatever it has to be. Under about half an inch of offset nobody sees it at all. Between half an inch and two inches it reads as a mistake, and that is worth correcting with fillers or a retyped cabinet before anything is ordered.
The faucet. A counter is 36 inches off the floor and a sill above a run usually lands between 40 and 48 inches, so the bowl and the cabinet are never in the same field of view as the glass. The faucet is a vertical object 12 to 18 inches tall standing right in front of the opening at eye level, and it is what a photograph of that wall shows. The cabinet joint below is an eighth of an inch reveal at knee height that nobody has ever noticed.
Double the offset, because the eye compares the two gaps either side of the faucet rather than the centre line. A half inch off gives a one inch difference between the gaps, which is invisible on a drywall return and findable against a casing. One inch off gives two inches and most people read it as wrong once it is pointed out. Two inches off gives four and it is obviously a mistake.
Rebalance the fillers at the two ends. Base fillers come as a 3 inch and a 6 inch board and both are ripped down on site, so making one narrower and the other wider by the same amount slides the whole run sideways with nothing else changing. The limit is the filler that has to grow, because a base filler stops at 6 inches, and the practical floor is about half an inch because anything thinner will not scribe to a wavy wall.
Yes. Catalogue widths step in 3 inch increments from 12 to 36 inches and then jump to 42, so there is no 39 inch box, and typing any width you like onto a cabinet is allowed in the planner, which then marks that line as resized on the quote. Choose a box that has no identical twin beside it, because what actually shows is a door two inches narrower than the one next to it. Never retype one of a matched pair flanking a range, and never retype the sink base you already sized to the bowl.
Further than most people expect. Under about 3 inches nothing changes at all, because the trap arm and the flexible supply connectors absorb it inside the cabinet. Out to roughly a foot, a longer trap arm usually does it, provided it keeps falling at about a quarter inch per foot and stays within the length allowed before the fixture needs its own vent. Past that the stub out has to move, which means opening one bay of wall. Trap arm lengths and venting vary by jurisdiction, so confirm with a licensed plumber.
Centre on the strongest pair of vertical lines in the assembly. On a trimmed window that is the outside edges of the casing, because those are the two lines the eye brackets the faucet between. On a drywall returned opening it is the two return corners, and on a large fixed pane in a slim frame it is the glass itself. The two centres are often different numbers, especially where a casing was ripped narrower on one side to clear a corner, or where a slider has a fixed half and a sliding half of different widths.
Then centring the sink on the window puts the sink off centre in the run, which is fine, because there is no line to compare a sink against on a run. The real constraint is landing space: a sink needs clear counter on both sides and generously on the side dirty dishes arrive. If chasing the window leaves under about 18 inches on one side, the window loses. Centre on the run instead and shift far enough that it reads as a decision rather than a near miss.
Yes, and it is the standard trick. In a 33 inch equal double the two bowls are just under 15 inches each and the divider sits on the sink centre line, so shifting the faucet 4 inches to catch the window lands the stream about three and a half inches off the divider wall, in the inner quarter of the bowl, and it splashes. The same 33 inch sink as a 60/40 has a large bowl about 18 inches wide straddling the centre, so the same 4 inch shift lands comfortably inside it. Order the sink handed so the large bowl is on the side the faucet has to move toward.
Draw the wall, draw the run at real widths, look at it straight on in elevation, and measure the two gaps either side of the faucet before anybody orders a box.
Open the 3D planner →