UK driving test pass rate by month and region, where seasonality is real and where it isn’t

Section 1, The national pattern, revisited

The national series for 2022-23 to 2024-25 sits between 47.9% in May and 49.2% in December. A 1.3-percentage-point swing across the calendar year, real and reliable, but small compared with the 35-plus percentage-point spread between UK centres or the eight-point spread between UK regions. The chart below is the same data shown on the original seasonality research page for cross-reference. The interesting question this page asks is whether the December peak holds up when you split the national figure regionally.

The national series has a December high, a January-to-March plateau slightly above the annual average, an April-to-June dip that bottoms out in May, and a July-to-November recovery. That national shape is the figure most learners already know. What the per-region split below shows is that this national shape is essentially the volume-weighted average of eleven regional shapes that mostly do not look the same as the national one.

Section 2, Three GB countries, three different shapes

The most natural first cut is by country. England, Scotland and Wales each have a recognisably different month-to-month pattern, even though all three share the same DVSA examiner pipeline and the same DT1 marking sheet. Northern Ireland is missing because DRT122A does not carry NI centre data; the DVA equivalent is published separately and we do not currently ingest it.

England tracks the national pattern almost exactly because England is the bulk of the national sample. The April-June dip and the December-ish high are both visible. Best month August at 49.2%, worst month June at 47.7%, spread 1.5pp.

Scotland runs lower overall (47.2% volume-weighted across 140,976 tests in the window) and its calendar-month shape is messier. July is the high month at 49.0%; the dip sits in May at 46.2%. The Scottish series is sensitive to small island and Highland centres where a handful of extra passes in any one month can move the figure by several percentage points. The headline spread of 2.8pp is more than England’s, but the wider confidence bands mean the practical seasonal swing for a Scotland-based candidate is probably smaller than the chart suggests.

Wales is the highest-passing of the three GB countries by some margin (54.1% volume-weighted, against England’s 48.5% and Scotland’s 47.2%), and its calendar peak sits in October rather than December. The dip lands in January. The Welsh seasonal swing is 2.3pp. Welsh test centres skew more rural and coastal than English ones, and the October peak is loosely consistent with summer tourist traffic complicating routes in July and August before things quieten down for autumn. We would not over-claim that story; the Welsh sample is the smallest of the three and the shape could move with the next DVSA release.

Section 3, Where seasonality bites most, and least

Aggregating up to the eleven UK regions used by our wait-time and density research pages gives a finer-grained picture. The chart below shows each region’s best-month-minus-worst-month spread in percentage points. Larger bar means more seasonal, shorter bar means flatter year-round.

Yorkshire and the Humber tops the chart at 3.1pp between its best month (February) and its worst (September). That is more than double the national headline spread of 1.3pp. A learner whose local centre sits in Yorkshire and the Humbercan rationally treat the month they book as a real lever, though even at the top of this table we are talking about three percentage points of expected probability, not thirty.

South West sits at the bottom of the rankable list at 1.8pp. The seasonal swing exists but is small enough that month shopping is essentially a tie-breaker. The South Westaverage pass rate (49.6%) is roughly stable across the calendar year, which is the pattern a candidate should expect to see for most large English regions outside the spring booking surge.

The middle of the league is occupied by London (2.6pp), the West Midlands (2.6pp), and the East Midlands (2.6pp). All three sit close to the national-spread band of one to two and a half points. The pattern of which calendar month carries the peak is more varied than the swing magnitudes suggest.

Section 4, Regions where seasonality is reliable vs regions where it is noise

Not every regional row in the table above is equally trustworthy. A region needs enough monthly volume across the three-year window to make its month-to-month figures stable. Our threshold is 5,000 tests per month in the region’s quietest calendar month. Below that, a few hundred extra passes in any one month can move the headline by a full percentage point, and the seasonal pattern starts to dissolve into sampling noise.

The North East is the standout caveat. Its 2.9pp spread looks meaningful on paper, but several of its calendar months sit below 5,000 tests across the three-year window. A candidate in Tyne and Wear or Northumberland is probably better off taking the national pattern as their anchor and treating the regional caveat as a reminder that the per-month figures wobble.

Section 5, Which calendar months do regions actually agree on?

The national chart says December is the best month and May is the worst. Reading that as a booking heuristic assumes the regions agree. They mostly don’t.

Among the 10 regions that clear our volume threshold, December is the most-common best month, claimed by 3 of them. That is a plurality, not a majority. The remaining 7 regions identify another calendar month as their peak, which means a learner planning around "December is the best month" is right less than half the time at the regional level.

The worst-month picture is slightly cleaner. The most-common worst month is June, claimed by 3 of the same significant regions. That is consistent with the national pattern’s May dip and aligns with the spring-booking-surge hypothesis: when test volume surges in late spring, the marginal new candidate is on average less prepared, and the volume-weighted pass rate drifts down. The pattern repeats across most regions because the booking surge is national.

The honest takeaway: the "best" month is regionally unstable. The "worst" month is regionally consistent. A candidate avoiding the late-spring dip is on safer ground than a candidate trying to time the December peak.

Section 6, Why regions show different seasonal shapes

We can describe the per-region variation. Explaining it is harder because DVSA does not publish the per-region operational data that would let us pin down the mechanism. Four plausible drivers, each labelled honestly.

Driver 1, weather sensitivity (probable, partially supported)

Rural-route centres in Scotland, Wales and the south west see more weather-sensitive driving conditions than central London ones. Single-track A-roads, livestock crossings, exposed coastal sections: these are environments where a damp October morning materially changes the driving task. The seasonal swing in rural-leaning regions could plausibly track local weather patterns more closely than the urban regions. Wales peaking in October and dipping in mid-summer is consistent with this; the West Midlands dipping in June is less so.

Driver 2, school-holiday cycles (probable, sample-limited)

UK school summer holidays run roughly July through August in most of England and Wales but late June through mid-August in Scotland. Candidates aged 17 to 18 are a meaningful slice of new test bookings, and they cluster their lesson volume in holiday weeks. The Scottish peak in July, English peaks scattered between January and August, and Wales’s October peak are loosely consistent with the regions’ different summer-holiday windows shifting the candidate-pool mix. We cannot prove it from DRT122A because the release does not split pass rates by candidate age.

Driver 3, examiner staffing (speculative)

DVSA examiner staffing varies by centre and by month. Annual leave clusters around school holidays in centres serving commuter towns; recruitment cycles affect different regions differently. Newly-deployed examiners on average mark slightly stricter while they settle into a centre, which could shift the local pass rate by a fraction of a percentage point. We have no way to confirm or deny this from the public dataset. A Freedom of Information request to DVSA could clarify per-region examiner-mix data, which would be a useful direction for a follow-up post.

Driver 4, route-mix changes (weak)

DVSA periodically updates the published route list at each centre. A new route introduced at a centre can shift the average difficulty of that centre’s test by a small margin, and if the update lands in a particular month the volume-weighted pass rate for that calendar month at that centre can drift. Aggregated up to the regional level, this is almost certainly too small an effect to drive the patterns we see, but it is the kind of thing that would need a per-centre per-month panel to rule out cleanly.

Section 7, What this means for a learner choosing a booking month

The seasonal effect is real and the effect is small. Three things worth keeping clear before reading any month-shopping strategy.

The regional best month matters more than the national one. The national figure is the volume-weighted average of eleven regional shapes that mostly do not peak in the same month. A learner in London should treat London’s peak (January at 49.5%) as more relevant than the national December figure. A learner in Wales should look at Wales’s October peak. The per-region table in section 4 above is the practical lookup for this.

The late-spring dip is the more reliable signal. 3 of the 10 significant regions identify a late-spring month as their worst. If the choice is between "book in May" and "book later in the summer", the data is reasonably confident that later in the summer is better, almost everywhere.

Centre and preparation dwarf everything else. Two candidates booking in the same month at different centres can face a 35-plus percentage-point gap in expected outcome. The seasonal effect at any single centre is one to three points. Use the easiest centres / hardest centres rankings to decide where to book. Use this page to decide when, treating it as a tiebreaker rather than a strategy.

Section 8, Methodology and limitations

Data source. DVSA DRT122A, three financial years pooled (2022-23 to 2024-25). Each row in the source gives a centre-month-gender combination with test count and pass count. We aggregate to the all-gender total per centre per month, then bucket centres into UK regions using the same heuristic as our wait-time research page, and finally sum across centres within each region to produce volume-weighted per-month figures. Licensed under the Open Government Licence v3.0.

Region bucketing. Each centre is assigned to one of eleven regions (London, South East, East of England, West Midlands, East Midlands, North West, North East, Yorkshire and the Humber, South West, Wales, Scotland) based on its OSM geoDisplayName. The bucket set is identical to the one used on /research/wait-time-by-region so a reader cross-referencing can use the same set. Northern Ireland centres are excluded because DRT122A does not cover them.

Significance threshold. A region is labelled "significant" when every calendar month in its three-year series carries at least 5,000 tests and its best-to-worst spread is at least 1.0 percentage point. 5,000 tests per month is the floor below which a single percentage point of seasonal swing is hard to distinguish from sampling noise; the 1.0pp swing floor keeps regions that are essentially flat year-round (where any calendar pattern is indistinguishable from random month-to-month variation) out of the headline comparison. Of 11 regions included, 10 clear both thresholds.

Volume-weighted aggregation. Per-region per-month pass rate is sum of passes across all centres in the region in that calendar month, divided by sum of tests. This is the same convention DVSA uses for its own quarterly press release headlines, and the same convention used throughout PassRates.uk. Documented on the methodology page.

Redacted rows. DVSA redacts centre-month cells with fewer than five tests to ".." under their disclosure-control policy. We skip these rows in the aggregation because they carry no usable count to weight by. Across the three-year window, redacted rows account for under 0.1% of test volume, and they concentrate at small island and Highland centres which is one reason Scotland’s monthly shape should be read with wider confidence bands than England’s.

What this analysis does not do. We do not adjust for COVID disruption (the three-year window starts in 2022-23, after the immediate post-lockdown booking surge, but the lingering effect on candidate mix is non-zero). We do not split by candidate age, by first attempt vs retake, or by examiner. We do not include Northern Ireland. We do not cover motorcycle or heavy-vehicle tests, which DVSA publishes separately. Each of these is a future research direction. Email hello@passrates.uk if you spot a mistake.

Cite this page: passrates.uk research/pass-rate-by-month-and-region v1.0 (2026). Data: DVSA DRT122A "Driving test and theory test data, cars" for 2022-23 to 2024-25, OGL v3.0. Aggregation: 287 UK car centres, 1,774,270 monthly tests, volume-weighted within 11 regions.

Quick-reference: best and worst month by region

Frequently asked questions

Related on PassRates.uk

• How we calculate pass rates
• Seasonality of UK pass rates
• Wait times by region
• London vs UK pass rate
• UK driving test volume trends 2019 to 2025
• Centre volume vs pass rate
• Pass rate vs population density
• Easiest UK test centres
• Hardest UK test centres
• Should I travel for an easier test centre?