Start with the vocabulary rather than the advice, because the vocabulary is the finding.
Across all 57 size charts we hold — read out of the images one seller published beside their items, and shown per item on the shelf — exactly seven measurement names appear. Length, on all 57. Waist on 33. Hip and leg opening on 32 each. Chest on 25, shoulder on 25, sleeve on 23.
That is the complete list. No neck. No collar. No cuff, no wrist, no armhole, no yoke, no back length taken separately from front.
A dress shirt at retail is sold on two numbers, and the first is the collar. In this corpus that number does not exist.
What a shirt is actually sized on here
Twenty-three of the 57 charts carry the four-field upper-body shape: length, chest, shoulder and sleeve. Anything shirt-like is sized on those four and nothing else, and three of them do the work.
Chest is the field that can still separate two sizes. Nineteen of the 55 adjacent chest steps in the corpus exceed the seller's own stated one-to-three-centimetre tolerance, which is the highest rate of any upper-body field.
Shoulder is the field that decides whether the shirt reads as tailored or as borrowed. Fifty of 55 adjacent shoulder steps sit inside that same tolerance, so the row will rarely choose between two neighbouring sizes for you — but a shoulder seam landing off your actual shoulder point is the single most visible fit error on a collared shirt, and it is not something an alteration on a replica is worth paying for.
Sleeve is close to inert. The median step between one size and the next is a single centimetre, the largest anywhere in the set is three, and across a whole size run sleeve moves a median of two centimetres. The full arithmetic is here, and its consequence for shirts is blunt: you cannot buy a shirt sleeve length off these charts the way you would order a 34/35 at retail. The grid does not have that resolution.
Sleeve figures across the corpus also run from 22.5 to 75 centimetres, because short-sleeved and long-sleeved garments share the field name with nothing to distinguish them. Only one chart describes a run where every sleeve figure is under 35 centimetres; the other 22 start at 48. If a sleeve number looks impossible, you are probably looking at a different garment than you pictured.
The collar, and what to do about its absence
The collar is not a detail you can size around. It sets where the shirt sits on the neck, whether it closes at all, and how it behaves under a jacket. Retail sells shirts on it for a reason.
Since no chart here publishes it, there are two places it can come from. The first is the seller, before you order. The message that gets answered is short and names the fields, and the full version of that request is written up separately. For a shirt, ask for the collar measured flat along the band with the top button done up, the cuff opening measured flat, and confirmation of which size those came from. An answer without a size attached to it is not an answer.
The second is your agent, once the item is in their warehouse. That is the only measurement in the chain taken from the object you are actually buying, and asking for a tape across the collar band costs nothing. Reading QC photos covers what else to request while you have their attention.
Formalwear raises the cost of every error
Three things make a collared shirt or a suit less forgiving than the rest of a haul.
- It is worn buttoned. A hoodie that is snug simply fits differently. A shirt that is snug at the chest gapes between the buttons, and no version of that reads as intentional.
- It is worn layered. A shirt sized for wearing alone and one sized for wearing under a jacket are different chest figures on the same body. Decide which you are buying before you open the chart, then measure a reference garment in that configuration — the same discipline that applies to sizing outerwear.
- It is usually bought for a date. Formalwear is ordered against an occasion, which removes the option of accepting the item, learning from it and buying better next time. If there is a date attached, build the timeline first — ordering for a deadline covers how much slack that needs.
Build the reference from a shirt you own
The method is the same as everywhere else on this site and it is the only one that survives a missing chart.
- Take a collared shirt you own that fits correctly. Lay it flat, buttoned, seams straight, no tension on the fabric.
- Measure chest armpit seam to armpit seam, shoulder seam to seam across the back, sleeve from shoulder seam to cuff edge, and length from the high point of the shoulder to the hem. The flat-measure walkthrough covers the landmarks properly.
- Measure the collar band separately, flat, buttoned, end to end. You now hold the number no chart in this corpus prints.
- Rank the fields before you compare: shoulder as a veto, chest as the decision, length for where the hem sits, sleeve as a description rather than a choice.
What we do not know, and one thing we will not tell you
The 57 charts come from a single store, MVT, and nobody here has bought one of these shirts and re-measured it. We publish the numbers the seller printed with a link back to the image — a record of a claim rather than a verification of one, for the reasons on the methodology page.
We also hold no quoted buyer account of how a replica dress shirt fitted. The quote corpus behind this site is largely hoodies, tech fleece, outerwear and sneakers. On rep formalwear fit specifically the record is empty, and an empty record is worth more said out loud than filled in with a guess.
And the question that always attaches itself to formalwear — whether anyone at the table will be able to tell — is not one this site answers. These are replicas. Nothing here will describe an item as indistinguishable from an original, and no fit advice above should be read as a claim about what a garment will pass as. Buy it because the shoulder measurement matches yours.