Age grading for cycling
How We Calculate It and What It Means
If you’ve looked at your triathlon results on the Serpentine website, or at the results for a bike race, you may have noticed an “age grade” percentage next to your bike split. This post explains what that number means, how we now calculate it, and why we’ve updated the method.
What is age grading?
Cycling — like running — is a sport where performance naturally changes with age. A 65-year-old completing a 40km time trial in 58 minutes is doing something harder and more unusual than a 30-year-old doing the same time. Age grading is a way of putting both performances on the same scale, so you can ask: how good was that ride, accounting for the athlete’s age?
The idea is simple. For every distance, every age, and for men and women, we hold an estimated “standard” — our best estimate of the fastest time a person of that age could theoretically ride that distance. Your age-grading percentage is then:
Age grade % = (standard time ÷ your time) × 100
So if the standard for a 65-year-old man over 40km is 3,200 seconds, and you rode it in 4,000 seconds, your age grade is 80%. A higher percentage means a better performance relative to what is possible at your age.
For running, where age-grading is quite common, the conventional wisdom is that scores above 90% are world-class, above 80% are national-level, above 70% are strong regional standard, and most recreational athletes fall somewhere in the 50–70% range. We don’t yet know if the same will be true for cycling — looking at the results so far the cycling percentages are roughly comparable. The important thing is that the percentage lets you compare performances across ages — and track your own improvement over time even as the raw times naturally get a little slower.
For example, here are the standard times for each age for 90km.
Standard times for 90km for each age
90km cycling age-grading standard. Hover over the chart to see speed and time at each age.
How we built the new standards
The standard at any given age is built in two parts: a baseline for peak-age athletes, and a curve showing how performance changes across the lifespan.
The baseline — world best times
For peak-age athletes (roughly 19–39, for reasons explained below), we use the world's best known open-road time trial performances at each distance. These come from the CTT (Cycling Time Trials) all-time records, which is the authoritative archive of British — and effectively world — road time trial records. The key anchors we use are:
For men: Marcin Bialoblocki's 42:58 for 25 miles (2018, 56.2 km/h) and John Archibald's 1:30:08 for 50 miles (2025, 53.6 km/h).
For women: Hayley Simmonds' 49:28 for 25 miles (2016, 48.8 km/h) and her 1:42:20 for 50 miles (2016, 47.2 km/h). For 100 miles, we use Joanna Patterson's 3:36:31 (2022).
This means that 100% represents a genuine world-record-level performance — consistent with how running age grading works.
Often the distances we ride, such as in a triathlon, don't correspond exactly to the distances for time trials, so we use log-linear interpolation across the known data points to derive standards at each target distance.
The age curve — VTTA standards
For athletes aged 40 and over, we use the age-degradation curve from the Veterans Time Trial Association (VTTA) 2025 standards, compiled by Jon Fairclough from analysis of approximately 11,000 elite solo cycling results. A big thanks to Hilary Walker for putting me on to these. The VTTA curve tells us, for each age from 40 to 95, how much slower an athlete of that age would typically be compared to a 40-year-old. We apply that curve on top of our world-best baseline.
For younger athletes aged 12–18, the VTTA published junior adjustments in 2024 derived from real CTT race data — so those age groups also use cycling-specific evidence rather than guesswork.
Why the standard is flat from age 19 to 39
You might expect peak performance to occur around age 25–28 and then gradually decline — that's roughly what happens in running. Cycling turns out to be different.
When the VTTA analysed their large dataset of competitive cyclists across all ages, they found no statistically meaningful age effect for open riders between 19 and 39. The performance of elite cyclists in this age band is essentially flat — unlike running, where a 35-year-old is measurably slower than a 27-year-old at world-class level.
This makes sense: cycling is less weight-bearing and less biomechanically demanding than running, and the aerobic and muscular systems that matter most in time trialling remain close to their peak for a longer stretch of adult life.
In practice, this means every athlete aged 19–39 is measured against exactly the same world-best standard.
These are time trial standards — triathlon splits will normally score lower
One important thing to understand: these standards are calibrated for standalone time trials, where a cyclist goes as hard as possible over the whole distance with no other demands on the body.
In a triathlon, you've already swum before getting on the bike, and you know you have a run to come. Even the most disciplined athlete will hold something in reserve.
This means your age grade for a triathlon non-drafting bike split will naturally be somewhat lower than it would be if you'd ridden the same course as a standalone TT — and that's correct, not a flaw. The number is telling you how that bike leg compares to the very best solo effort possible, which is a meaningful benchmark even if it's one you're not expected to reach in a triathlon context.
In draft-legal racing riders may benefit from riding in a group, which can push bike splits faster than a solo effort would achieve.
This is also the approach to age-grading the run leg of a triathlon. We use the world best time in a stand-alone race - so we'd expect the best triathletes to be a bit slower than that on their run leg.
What this replaces
The previous version of this calculation for cycling used a different approach: rather than published time trial standards, we looked at the fastest bike splits recorded across a range of triathlon races, and used those as our age-group benchmarks. The idea was reasonable — use the best observed performances in triathlon — but we didn't have enough data to construct reliable standards. A single exceptional split in one race (perhaps on a fast, point-to-point course, or in a draft-legal event, or where the course was short, or errors in results) could skew the standard for an entire age group. Some age groups ended up with standards that were implausibly fast; others had standards that were too easy.
We do have some club members - notably Hilary Briggs - who genuinely are riding the fastest bike splits in the world in their age group. But we were getting implausibly high age grading figures for many races, which suggests that standards constructed this way were not fit for purpose.
The new approach is on a firmer foundation: it is grounded in carefully verified record performances and a well-evidenced age curve built from thousands of results, with the same philosophical basis as the running age grading you'll already be familiar with from your 5k and parkrun results.