How to measure drop-handlebar width: hoods, drops and outside edges
Name the bar station and centre/edge convention before comparing road or gravel widths. Repeat a real span without turning shoulder breadth into an ideal size.
Published 9 October 2026 · OpenBikeFit
Control reference field guide · not to scale
Same sections. Different endpoints.
Fictional bare circular sections: diameter 24 mm each.
Centre span 400 mm · outside 424 mm · inside 376 mm.
Conditional construction only; not a universal conversion or ideal bar width.
On this page
- 01Two definitions belong beside every width
- 02Read the exact manufacturer convention
- 03A conditional centre-to-edge example
- 04Keep upper and lower stations separate
- 05Repeat the observation independently
- 06Shoulder breadth is another quantity
- 07Preserve the record before choosing a next step
- 08What this comparison cannot settle
Choose the quantity first
Inspect the width and resting-gap workflow
Existing free tools keep their access and hardware boundaries. Reading a guide does not create a bike, activate a sensor or approve work.
Open toolA road or gravel handlebar can have a narrow upper span and wider lower ends while carrying just one size on the label. A second bar can use the same printed size but measure from outside edges instead of tube centres. Before comparing those labels or entering a tool, write what each number actually joins. The useful first result is a named measurement, not a verdict that the bar is too wide or too narrow.
This field guide concerns the assembled bicycle and non-destructive observations. Do not remove tape, loosen controls or rotate a bar merely to obtain an easier number. If the relevant centres are hidden, retain that uncertainty and consult the exact drawing. A catalogue dimension, an installed measurement and a rider’s hand position are three different records; preserve their identities instead of treating one as a replacement for another.
Two definitions belong beside every width
The station is where along the bar the span is taken: a named control-clamp location, a named part of the drops or a specified pair of bar-end points. The convention says whether endpoints are centres, outer edges or some other marked features. Record both in words. Width at the drops is incomplete if the selected height or fore-aft station can move along a curved tube.
For example, write hood-clamp centres at the declared clamp station rather than simply hoods. A mounted hood body may project inward or outward from its clamp, and its rubber cover is not a tube centre. A widest visible span is a different task again. It cannot silently replace the upper span used in a component drawing, even if it is easier to reach with a ruler.
Read the exact manufacturer convention
Deda’s Superzero Gravel page explicitly describes its size at the lever position using outside-to-outside endpoints and separately lists full width. Zipp’s SL 80 Race discussion uses centre-to-centre size naming. These are examples of why a naked size label is insufficient, not a claim that every product from either brand shares one permanent convention. Retain model, size, drawing and source date with the number.
Do not compare brands by subtracting a fixed universal amount from every outside width. The local tube shape, measurement direction, selected station and coverings matter. A drawing can resolve one manufactured convention without telling you exactly where your installed control or hand sits. Manufacturer marketing about control, comfort or aerodynamics is not adopted here as a measured outcome for the reader’s bicycle.
A conditional centre-to-edge example
Imagine two equal circular tube sections measured along a line through both centres, with no tape or other covering. Their fictional centre spacing is 400 mm and each section diameter along that line is 24 mm. Outside spacing is 400 + 12 + 12 = 424 mm; inside spacing is 400 − 12 − 12 = 376 mm. The same unchanged pair has three different valid spans.
That construction explains the reference difference; it is not a conversion rule for an assembled curved handlebar. Unequal or non-circular sections need their actual endpoint offsets along the measurement line. Tape changes an observed edge and may hide the original tube boundary. Record the measured feature directly rather than labelling a guessed centre as observed. Displaying tenths of a millimetre would not repair an unresolved endpoint.
Keep upper and lower stations separate
A fictional bar has 400 mm between named upper centres and 460 mm between named lower centres. The lower station is therefore 60 mm wider overall. Under an explicitly symmetric construction, each side is 30 mm farther outward. Neither the 60 mm difference nor the 30 mm half-span is the bar’s flare angle. They describe distances at these particular stations.
Ritchey’s gravel guide treats bar size and shape as separate descriptions and explains why flared drops can lead to two widths. For your record, draw the chosen stations rather than assuming that all points called drops are identical. If the left and right features are not mirror locations, the subtraction includes a reference mismatch. Move the measuring tool, not the bicycle parts, to resolve it.
Repeat the observation independently
Stabilise the bicycle under one stated alignment, mark non-damaging endpoints and use a suitable ruler or tape along the declared span. Remove and reposition the measuring tool before the second reading. Keep the raw values and method. Reading a ruler twice while leaving it in the same ambiguous position tests less of the process than independently locating the endpoints again.
The existing width assistant uses a 5 mm pair-agreement check and requests a fresh third reading when needed. That is a product convention, not instrument accuracy, anatomical precision or a confidence interval. Do not use its threshold to declare a part safe. If the station changes between readings, start a separately named series instead of averaging different widths into one convenient number.
Shoulder breadth is another quantity
Marked bony shoulder points and outer soft-tissue edges describe different body spans. Neither is automatically the same kind of endpoint as a bar edge. The assistant keeps those definitions explicit and does not mix centre and outer-edge conventions. A measurement should stay within the method that produced it; do not change the definition merely to make the body and component values agree.
For fictional repeated centre-based observations, shoulder readings 410 and 412 mm average 411 mm; upper bar readings 420 and 422 mm average 421 mm. Bar minus shoulder is +10 mm. This reports the observed relationship under those definitions. It supplies no instruction to buy a 411 mm bar, narrow the current bar or treat equality as improved fit, comfort or handling.
Preserve the record before choosing a next step
Keep component identity, catalogue label, actual station, endpoint convention, independent raw readings, coverings and the alignment method together. Mark unknown fields as unknown rather than zero. A small sketch can be more useful than a photograph without a scale or reference. If a second bar is compared, give it its own complete declaration rather than copying the first bar’s convention.
Use the free width tool for its observed repeated spans or the blank worksheet to retain the protocol. Read the shape guide when the lower width differs; use the separate control guide when the question is a resting blade gap. These pages neither select a bar nor approve an assembly change. Exact current manuals and physical checks remain necessary before any separately justified mechanical work.
What this comparison cannot settle
A wider measured lower span does not prove better stability, a narrower upper span does not prove an aerodynamic benefit, and a shoulder match does not establish a suitable riding position. Those are different questions requiring different evidence. A completed dimension record is useful because it prevents mistaken equivalence, not because it converts the ruler into a test of rider outcome.
Leave the installed bicycle unchanged while collecting the baseline. Stop if a measurement cannot be made without stressing a control, hose or component. A qualified mechanic can resolve assembly and function questions; the browser cannot inspect them. Reading a guide requires no bike record, camera or sensor, and following a tool link retains that tool’s existing consent and hardware boundaries.
Practical questions
Frequently asked questions
Is drop-bar width measured at the hoods or the drops?
It depends on the exact product drawing and the task. Record the station and whether endpoints are centres or edges; keep mounted hood-body spans separate from bar-clamp-centre spans.
Can I subtract 24 mm from every outside width?
No. The fictional circular-section example needs equal bare sections and a measurement line through both centres. Real shape, station, angle and coverings can invalidate that conversion.
Should my handlebar width equal my shoulder width?
An observed difference does not prescribe equality or an ideal size. Preserve both measurement definitions; comfort, handling and suitability remain separate questions.
Do two identical size labels mean identical widths?
Not unless their station and endpoint conventions also match. Compare the exact model drawings and installed observations rather than assuming label equivalence.