Tyre Dispatch · Workshop Data
Tyre Pressures and Inflation Wear
Every worn tyre carries a physical record of the pressure it ran at. We read that record across 1,149 tyres. It shows that under-inflation dominates ordinary cars - and reverses completely on utes, vans and light-truck tyres, where the fault is the opposite one.
A correctly inflated tyre wears roughly evenly across its width. Run it soft and the contact patch bows outward, so the shoulders carry the load and wear first. Run it hard and the centre bulges into the road and wears first. The tyre keeps that record until it is scrapped.
We take tread depth at four positions across the width of most tyres through the workshop, which is enough to recover the shape. Both shoulders low by a similar amount means it ran soft. Centre low means it ran hard. If only one shoulder is low, pressure is not the explanation - that is a wheel-alignment signature, and we hold it separately throughout rather than let it contaminate the inflation figures.
Figure 1 · Wear shape across 1,149 worn tyres
Just over four tyres in ten carry a readable pressure signature - 476 of 1,149. Under-inflation outnumbers over-inflation 1.8 to 1, which makes physical sense: tyres lose pressure by being ignored, whereas over-inflating one takes an active decision.
That is the fleet-wide picture, and on its own it is the answer you would expect. It is also misleading, because it averages together two populations that behave in opposite ways.
FINDING 01The pattern inverts on high-pressure tyres
Split the same tyres by what the sidewall is rated to hold, and the fault reverses.
Figure 2 · Inflation fault by tyre construction
| Group | n | Median max sidewall PSI |
Under- inflated |
Over- inflated |
|---|---|---|---|---|
| Standard passenger | 1,047 | 50 | 28.5% | 13.0% |
| LT - light truck | 77 | 68 | 6.5% | 40.3% |
| Dual load index (e.g. 121/120) | 92 | 73 | 7.6% | 28.3% |
On ordinary passenger tyres, under-inflation beats over-inflation more than two to one. On light-truck tyres, that flips to six to one in the other direction. Dual-load tyres behave the same way.
The explanation is load, not carelessness. A tyre rated to 68 or 73 psi is designed to carry weight, and the pressure that suits it loaded is far too high for it empty. Most utes and vans spend most of their lives empty. The tyre is at the pressure the placard asks for, and it is still over-inflated for what it is actually carrying - so the centre wears first.
Two more cuts of the same data show the same reversal, and both are gradients rather than categories - which is harder to explain away as a quirk of one grouping.
Figure 3 · The reversal as a gradient - section width and speed rating
| Section width | n | Under | Over |
|---|---|---|---|
| 155 – 175 mm | 118 | 36.4% | 4.2% |
| 185 – 205 mm | 310 | 31.0% | 7.1% |
| 215 – 235 mm | 414 | 27.1% | 12.6% |
| 245 – 265 mm | 252 | 19.8% | 28.2% |
| Speed rating | n | Under | Over |
|---|---|---|---|
| H - 210 km/h | 347 | 36.9% | 9.2% |
| T - 190 km/h | 128 | 32.8% | 10.2% |
| S - 180 km/h | 56 | 25.0% | 25.0% |
| R - 170 km/h | 40 | 10.0% | 32.5% |
| Q - 160 km/h | 31 | 0.0% | 58.1% |
Under-inflation falls steadily as tyres get wider while over-inflation rises, crossing over around 245 mm. Speed rating tells it more starkly still: not one Q-rated tyre in the set showed under-inflation, and 58.1% showed the opposite.
The terrain data points the same way. All-terrain tyres - mostly on 4WDs run at moderate pressure - show the ordinary under-inflation bias. Mud-terrains, which sit on the heaviest and highest-pressure setups, are the most centre-worn group in the dataset.
Figure 4 · Inflation fault by terrain category
| Terrain | n | Under | Over | Alignment | No pattern |
|---|---|---|---|---|---|
| All terrain (AT) | 86 | 31.4% | 22.1% | 14.0% | 32.6% |
| Highway terrain (HT) | 80 | 16.2% | 21.2% | 7.5% | 55.0% |
| Mud terrain (MT) | 19 | 0.0% | 57.9% | 15.8% | 26.3% |
FINDING 02Profile, grade and pattern each shift the odds
Figure 5 · Fault mix by sidewall profile
| Aspect ratio | n | Under | Over | Alignment |
|---|---|---|---|---|
| Ultra-low - 45 and below | 303 | 25.1% | 15.2% | 20.8% |
| Low - 50 to 55 | 250 | 29.6% | 12.0% | 6.8% |
| Mid - 60 to 65 | 475 | 32.8% | 12.0% | 11.2% |
| High - 70 and above | 152 | 15.8% | 21.7% | 15.1% |
The clearest profile finding is not about pressure at all: ultra-low profile tyres carry three times the alignment-wear rate of low-profile ones - 20.8% against 6.8%. Short stiff sidewalls transmit a camber or toe error straight into the tread instead of absorbing it, and they sit on cars whose suspension geometry is more aggressive to begin with.
The same signal shows in tread pattern. Asymmetric patterns - overwhelmingly a performance fitment - show 17.6% alignment wear against 11.0% for symmetric patterns. Directional patterns sit lowest of all at 6.2%, though on only 48 records.
Figure 6 · Fault mix by UTQG treadwear grade
| UTQG treadwear | n | Under | Over | Alignment |
|---|---|---|---|---|
| Below 300 - soft, performance | 133 | 12.8% | 13.5% | 25.6% |
| 300 – 399 | 232 | 25.4% | 19.0% | 12.9% |
| 400 – 499 | 333 | 43.2% | 4.5% | 9.6% |
| 500 – 599 | 136 | 33.1% | 14.0% | 9.6% |
| 600 and above - hard | 59 | 28.8% | 22.0% | 6.8% |
Soft performance compounds graded under 300 show the highest alignment wear of any group at 25.6%, and the lowest under-inflation. Mid-grade touring rubber at 400–499 shows the opposite extreme: 43.2% under-inflated and only 4.5% over, the most lopsided ratio in the dataset. These are ordinary tyres on ordinary cars that nobody checks.
UTQG temperature grade separates the two far less: grade A runs 32.3% under and 11.3% over, grade B runs 24.1% and 20.7%. The direction is consistent with grade B appearing more often on load-rated fitments, so we do not read it as a temperature effect.
FINDING 03The dangerous fault and the expensive fault are different faults
Alignment wear fails a check far more often than either pressure fault. But it is also rarer. Both facts matter, and quoting either alone gives the wrong answer.
Figure 7 · Failure rate against total failures produced
| Fault | Tyres | Share of fleet |
Fail rate | Failed tyres |
Share of all failures |
|---|---|---|---|---|---|
| Under-inflation | 306 | 26.6% | 46% | 141 | 25% |
| Over-inflation | 170 | 14.8% | 43% | 73 | 13% |
| Alignment (one side) | 159 | 13.8% | 70% | 112 | 20% |
| No distinct pattern | 514 | 44.7% | 44% | 228 | 41% |
| Pressure faults combined | 476 | 41.4% | 45% | 214 | 39% |
A tyre with an alignment fault is around half again as likely to fail as one with a pressure fault - 70% against 45%. But pressure faults are three times as common, so across the whole fleet they produce 1.9 times as many failed tyres: 214 against 112, and account for 39% of every failure we recorded against alignment's 20%.
FINDING 04The faults cluster the way physics says they should
Everything above rests on reading a wear shape and naming a cause. That inference deserves a test that does not depend on our own judgement, and the data contains one.
A pressure fault is a property of the vehicle: whoever neglects one tyre neglects all four, and whoever over-inflates does it round the car. A camber or toe fault belongs to one corner. So if our categories track something real, pressure faults should repeat across the tyres that came off the same car, and alignment faults should repeat much less. We grouped 669 tyres into 296 probable sets and asked exactly that.
Figure 8 · When one tyre in a set shows a fault, does another share it?
| Fault | Sets | Shared with another tyre | Rate |
|---|---|---|---|
| Over-inflation | 64 | 41 | 64% |
| Under-inflation | 118 | 63 | 53% |
| Alignment (one side) | 63 | 18 | 29% |
Pressure faults repeat across a set roughly twice as often as alignment faults do. That is what the physics demands, and it is not something our classifier was built to produce - had the categories been noise, all three would have shared at the same rate. They separate, and they separate in the predicted direction.
FINDING 05What it costs in rubber
When a tyre is legally finished because one part of the tread reached 1.5 mm while the rest still had depth, the difference is rubber bought and never used. We can add it up directly.
How we counted wasted tread - and why it is deliberately rough
Once any point on a tyre reaches 1.5 mm the tyre is finished, so we sum how much depth remains above 1.5 mm across its other grooves. A tyre reading 1.4 / 4.2 / 4.2 / 1.4 counts 5.4 mm wasted.
This overstates the true loss, and it is worth being explicit about why. An under-inflated tyre does not merely wear its shoulders - it also lifts the centre off the road, so the centre was never going to wear at the same rate. Take two identical tyres at 6 mm: run one soft until its shoulders hit the limit and it finishes near 1.4 / 4.2 / 4.2 / 1.4, an immediate fail with most of its middle intact; run the other correctly and the same total rubber consumed leaves it near 2.6 mm all round - still legal, still on the car. The honest comparison is between those two end states, not between 4.2 and 1.5. We use the simple sum because it needs no assumption about what would have happened, and we report it as an upper bound rather than a precise figure.
Figure 9 · Depth left above the limit on legally finished tyres
| Fault | Finished tyres |
Mean wasted per tyre |
Worst single tyre |
Total mm wasted |
|---|---|---|---|---|
| Alignment (one side) | 68 | 2.52 mm | 8.10 mm | 171.3 |
| Under-inflation | 57 | 1.93 mm | 4.90 mm | 110.1 |
| Over-inflation | 50 | 1.18 mm | 4.90 mm | 58.8 |
| No distinct pattern | 97 | 0.68 mm | 3.40 mm | 66.2 |
| Pressure faults combined | 107 | 1.58 mm | 4.90 mm | 168.9 |
Per tyre, alignment is comfortably the most wasteful fault: 2.52 mm against 1.58 mm for pressure, 60% more. The worst single tyre in the dataset was an alignment case that reached the limit on one edge while carrying 8.1 mm of usable depth across the rest of its tread - very nearly a whole new tyre thrown away.
Each of the three faults leaves a shape you can see without a gauge. Below is one tyre for each, photographed as it came off the car - all three picked because the inspector had flagged them as textbook cases in their notes at the time, before any of this analysis existed.
Figure 10 · One tyre for each fault, as it came off the car
But the arithmetic lands somewhere unexpected. Because pressure faults are so much more common, the total rubber wasted by each is almost exactly the same: 168.9 mm from pressure faults against 171.3 mm from alignment. Two very different failure modes - one rare and severe, one common and mild - cost the same in the end.
An even-wear tyre wastes 0.68 mm. That is the floor, and it is what all of this is measured against.
Those millimetre totals rank the faults against each other, but they are not a share of a tyre's life: they add four grooves together, so a tyre with four grooves each 1 mm clear of the limit scores 4 mm without having lost 4 mm of life. For a figure that can be compared with published estimates, we ask a different question - when one point forced the tyre off the car, how much usable depth was it still averaging?
Figure 11 · Share of usable service life left unused
| Fault | Finished tyres | Mean life unused | Worst case |
|---|---|---|---|
| Alignment (one side) | 68 | 7.6% | 30.4% |
| Under-inflation | 57 | 5.9% | 18.5% |
| Over-inflation | 50 | 3.4% | 18.7% |
| No distinct pattern | 97 | 2.1% | 13.1% |
| Pressure faults combined | 107 | 4.8% | 18.7% |
How this compares with the published figures
Industry estimates put the life lost to uneven wear at over a quarter of a tyre, and the share attributable specifically to under-inflation at 17.8%. Our measured figure is much lower - 5.9% for under-inflation, 7.6% for alignment.
We do not think this contradicts them, and we would not present it as a correction. Three differences explain most of the gap. Published figures are generally modelled across a tyre's whole service life, including the mileage penalty of running soft, whereas ours is a single measurement of depth remaining at the moment of removal. Ours counts only tyres that reached the legal limit, excluding those retired early with more depth left. And our four gauge positions never touch the shoulder, where the loss is greatest. Every one of those pushes our number down. It should be read as a hard floor on the waste, not an estimate of its true size.
FINDING 06Alignment faults kill tyres young
Using DOT date codes we can ask how old each tyre was when we saw it, and what share had already reached the limit.
Figure 12 · Age at inspection and share already finished
| Fault | n | Median age at inspection |
Already at or below 1.5 mm |
Mean spread across tread |
|---|---|---|---|---|
| Alignment (one side) | 157 | 4.5 yrs | 43% | 1.84 mm |
| Over-inflation | 171 | 4.5 yrs | 29% | 1.13 mm |
| Under-inflation | 331 | 4.3 yrs | 17% | 1.25 mm |
| No distinct pattern | 542 | 5.1 yrs | 18% | 0.57 mm |
The ages are similar across every group - around four and a half years. What differs sharply is how many were already finished by then. 43% of alignment-worn tyres had reached the limit, against 18% of evenly worn ones at a slightly younger median age. An alignment fault roughly doubles the chance a tyre is done by its fifth year.
Over-inflation sits in between at 29%. Under-inflation, at 17%, is no worse than even wear on this measure - its cost shows up in wasted rubber and in fuel, not in premature death.
The number on your sidewall is not the number you want
Nearly every tyre carries a maximum-pressure marking, and it is not really an independent fact about the tyre - it follows from how the casing is rated. Across our records it lands in tight clusters:
| Construction | Typical sidewall maximum |
|---|---|
| Standard load | 44 or 51 psi |
| XL / Extra Load | 50 psi |
| LT / light truck | 65 – 80 psi |
That figure is the maximum the casing may be inflated to. It is not the pressure the tyre should run at. The right pressure comes from the vehicle, not the tyre - it is on the placard in the driver's door jamb, and for most passenger cars sits between 30 and 36 psi. Inflating to the sidewall maximum is one of the most reliable ways to produce the centre wear in Figure 1, and on a light-truck tyre the gap between the two numbers is enormous: a sidewall reading 80 psi may want 35 unladen.
What we would actually do with this
- If you drive a normal car, you are probably soft. Under-inflation runs at 28.5% on passenger tyres and peaks at 43.2% on mid-grade touring rubber. Check monthly, cold, against the door placard.
- If you drive a ute or van, you are probably hard. The fault reverses on load-rated tyres - 40.3% over-inflated against 6.5% under. If you run empty most of the time, use the unladen figure on the placard, not the laden one, and only put the pressure up when you actually load it.
- One-sided wear is not a pressure problem. If one shoulder is going and the other is not, no amount of air will fix it. That tyre fails at 70%, wastes the most rubber of any fault, and the underlying fault will do the same thing to its replacement.
- Low profile and soft compounds deserve an alignment check. Ultra-low profile runs 20.8% alignment wear and sub-300 treadwear compounds 25.6% - both roughly double the fleet rate.
- Look across the tread, not down at it. A tyre with 5 mm in the middle and 1.4 mm on an edge is not legal, however healthy it looks from above.
Method, and what this cannot tell you
Records come from tyres documented at Tyre Dispatch between December 2025 and July 2026 - chiefly tyres presented for replacement, alongside new stock and inspection queries. Depth is taken at four positions across the width for most records and three for older ones.
This is not a prevalence survey, and the numbers here should not be read against ones that are. Published surveys measure pressure on vehicles in service - typically finding that most vehicles carry at least one under-inflated tyre, and that fewer than one in ten have all four correct. Those describe cars on the road on a given day. We describe tyres at the end of their lives, one tyre at a time, and we infer pressure from wear rather than measuring it. A tyre here represents a history, not a vehicle, and the two cannot be compared directly.
337 of the 1,545 records are excluded: those captured in a four-zone mode whose column ordering is not documented, those with incomplete readings, and an early block of 200 records captured before the measurement-mode field existed. We excluded them rather than guess, since a wrong assumption about column order would invent wear shapes that are not there.
We then tested that decision rather than assuming it. Re-running the whole classification with the 181 usable early records added moves almost nothing: under-inflation shifts from 26.6% to 26.8%, over-inflation 14.8% to 15.0%, alignment 13.8% to 15.3%. Had those early records used a different column order, scrambling shoulders and centre, the alignment share would have jumped sharply - it does not. That is good evidence the ordering convention has been stable since the first record, and it means the 1,149-record basis is conservative rather than biased.
A tyre is not moulded flat, and it matters
New tread is not the same depth right across the tyre. Measuring 59 genuinely unused tyres, the centre grooves are moulded deeper than the shoulders in 55 of them - by 0.49 mm on average for asymmetric patterns and 0.37 mm for symmetric ones. Asymmetric tread, which pairs a stiff shallow outer shoulder with deeper inner grooves, is the more lopsided by design.
That sits almost exactly on the 0.5 mm threshold used here, so an unused tyre can be classified as under-inflated purely on how it was manufactured. It showed up as an obviously false signal - 45.8% of unused tyres reading as under-inflated - and it is why every record graded as unused is now excluded from this report. A tyre that has not worn cannot carry a record of how it wore.
We also tested the obvious next step: subtracting each pattern's moulded profile from every tyre. We rejected it. Agreement with the inspectors' independent visual assessment fell from 85% to 51% for under-inflation while over-inflation held at 84% - the correction was erasing shoulder wear that a person had physically looked at and confirmed. Raising the threshold in stages did the same thing more gradually. On worn tyres the moulded offset is swamped by real wear in any case: symmetric and asymmetric patterns classify almost identically once a tyre has been used. The flat threshold stays, the unused tyres go, and the bias is disclosed here rather than silently corrected for.
Five limits worth stating plainly
- We never measured a pressure. Every inflation finding is inferred from wear geometry. It describes the pressure a tyre ran at across its life, not the pressure it arrived on.
- Wear shape has more than one cause. Sustained heavy loading, hard cornering and age also shape a tread. Symmetrical shoulder wear is most commonly under-inflation, but this dataset cannot prove causation in any individual case.
- The shoulder itself is not measured. Our gauge positions read the grooves. The outermost edge - where inflation and camber faults do their most visible damage - falls outside them, so every deviation reported here is conservative.
- Segments are not independent. LT construction, dual load index, high profile, mud terrain and high sidewall pressure describe substantially the same tyres. They are five views of one population, not five findings.
- Small groups stay small. Mud terrain is 19 records and directional patterns 48. We report them because the direction is interesting, not because they are settled.
One further caution on the size-level data. Splitting individual sizes by wear type leaves groups of three to twenty tyres, which is too thin to support claims about how long a particular size lasts. Suggestive patterns exist - 175/65R15 runs 50% under-inflated across 42 scans, and 195/65R15 under-inflated tyres were a median 2.4 years older at inspection than evenly worn ones - but we are not publishing them as findings until the counts support it.
More from the research programme
The Workshop Tyre Report: First 1,500 Inspections · Tyre Research Hub · Load Ratings and Load Capacity · Tread Pattern Direction
Tyre Dispatch · New Zealand · First edition · 29 July 2026
Descriptive analysis of one workshop's records. Not a representative sample of the national fleet, and
not a substitute for a warrant of fitness inspection. Legal minimum tread depth in New Zealand is
1.5 mm across the required portion of the tread. Corrections welcome:
email@tyredispatch.co.nz
Page © 2026 Tyre Dispatch NZ · Dataset, measurements and photographs © Taylor Houghton,
all rights reserved · Dataset licensing
