He'd shaved his position down over eighteen months — elbows tighter, torso lower, pads brought in another centimetre each winter until the fit guru called it "as low as your hips will allow." The wind-tunnel printout said his CdA had never been better. Race day told a different story: watts that wouldn't hold past twenty minutes, a power file that cratered exactly where it used to flatten out. He'd built the most aerodynamic position of his life, and it was the slowest he'd ridden that course in three years.

This isn't a training problem. It's a position problem, and it's one of the most measured trade-offs in cycling science — hiding in plain sight on every fit chart that chases a single number.

Section 01What chasing the lowest CdA number costs you

Nineteen well-trained male cyclists rode a cycle ergometer at five different torso angles — their own preferred position, plus four fixed angles at 0°, 8°, 16° and 24° from horizontal — while researchers tracked oxygen uptake, ventilation, gross efficiency, peak power output, heart rate and frontal area at each one (Fintelman 2015). As torso angle dropped toward horizontal, frontal area fell in a straight line, exactly as every fit guide promises.

Everything on the physiology side moved the other way. The maximal values reached in the incremental test — oxygen uptake, ventilation, gross efficiency and peak power output among them — decreased as torso angle dropped, and the differences were statistically significant across the board (P < 0.001) (Fintelman 2015). The 0° torso angle position — fully flat — significantly affected the metabolic and physiological variables compared to all other investigated positions, preferred angle included (Fintelman 2015).

It isn't only about your ceiling, either. At fixed submaximal loads — 60%, 70% and 80% of peak power, the intensity band a long time trial or a hard group effort actually lives in — the physiological cost of holding that same output climbed as torso angle dropped (P < 0.005) (Fintelman 2015). The lower position doesn't just cap what you can produce at your limit. It makes the exact effort you're already riding at cost more.

Position changes the drag side by a lot, too. Wind-tunnel testing of one rider in three positions — upright, dropped with straight arms, and a full time-trial tuck — measured frontal areas of 0.41 m², 0.37 m² and 0.34 m² respectively, a roughly 17% swing from most upright to most aggressive (Defraeye 2010). Nobody's arguing position doesn't move drag. The question is what it costs you to get there.

Section 02Why folding lower costs you power

Neither study set out to explain the mechanism directly — they measured the outcome, not the cause — but the pattern points somewhere obvious once you picture what's actually happening at 0°. Folding the torso down toward the top tube compresses the chest cavity and closes the hip angle at the same time. The diaphragm has less room to move, which is the textbook explanation for why ventilation gets harder exactly where these studies found it did. The tighter hip angle changes the leverage the big hip-extensor muscles — glutes and hamstrings — have on the pedal stroke, which is a plausible reason peak power specifically took a hit.

None of that is a new discovery on top of the two papers above — it's the standard biomechanical account for why trunk flexion costs endurance athletes output, and it fits the shape of what Fintelman's data actually shows: a steady, graded cost that scales with how far you fold, not a sudden cliff at one specific angle. You're not breaking a rule at 0°. You're compressing the engine to shrink the hood, and the engine notices.

It also explains why the submaximal finding matters more than the peak-power one for most riders. A restricted breath at your absolute limit costs you a number on a test you rarely repeat in anger. The same restriction sitting underneath every watt of a two-hour effort at threshold is a cost you pay the entire time you're racing — which is exactly the load Fintelman's team measured at 60%, 70% and 80% of peak power, not just at the ceiling (Fintelman 2015).

Section 03There's a speed-dependent sweet spot, not one "best" position

Fintelman's group followed up the next year with a modelling paper that mapped the trade-off across speed rather than testing fixed angles alone. Using the same rider data, two independent models — one minimising total metabolic energy cost, one maximising power output — converged on the same shape of answer: the optimal torso angle isn't fixed, and it's never fully horizontal (Fintelman 2014).

Aerodynamic losses only start to dominate the equation above roughly 46 km/h — squarely flat-TT and downhill-effort territory (Fintelman 2014). Below about 30 km/h — a long climb, a headwind grind, most of a hilly gran fondo — a more upright position actually wins once the physiological cost is priced in (Fintelman 2014). The lowest position on your fit chart is built for a speed a lot of your riding never reaches.

Section 04Test net speed, not the number on the CdA printout

The practical fix isn't abandoning aero positioning. It's testing the thing that actually decides races: sustainable power in the position you're planning to hold, not the drag number alone.

A simple version doesn't need a lab. Pick two or three torso angles you can actually hold — your current race position, one notch more upright, one notch lower if your flexibility allows it — and run the same steady effort in each: a 20-minute controlled interval at a fixed perceived effort, ideally on a stretch of road or trainer setting you can repeat. Compare average power, not just how each position felt in the moment — perceived effort and measured output don't always agree.

1. Ride the position you race in, not the position the wind tunnel liked best

   Book the fit session, then validate it on the road or a controlled trainer effort — sustainable power over 20+ minutes is the number that matters, not frontal area alone.
2. Match the position to the speed you'll actually hold

   The aero-versus-power optimum shifts with speed, favouring more upright below ~30 km/h and more aggressive above ~46 km/h (Fintelman 2014) — a flat 40 km TT and a hilly gran fondo spend very different amounts of time in each zone, so they rarely want the same torso angle.
3. Watch what happens at submax, not just at max

   At fixed submaximal loads, physiological cost rose as torso angle dropped — not only the maximal ceiling fell (Fintelman 2015). A position problem doesn't have to show up on a max test to be costing you.
4. Re-test position changes the way you'd re-test a new set of wheels

   Small torso-angle changes move frontal area in a straight line — treat every "lower" adjustment as a trade to verify, not a free win.

Section 05What to actually ask your fitter for

Most fit sessions are still built around one target: the lowest CdA the rider's flexibility will tolerate. That's the wrong brief if you race more than flat time trials. The modelling data puts the crossover at roughly 30 and 46 km/h — aerodynamics dominating above the higher figure, a more upright position winning below the lower one (Fintelman 2014). Ask for two positions, not one: a full aero tuck built for the speeds you'll actually hold in a flat TT, and a slightly more open position for anything hillier or slower, where a full tuck spends more time above your physiological break-even than below it.

Ask the fitter to check submaximal comfort, not just maximal squeeze. At fixed submaximal loads, physiological cost rose as torso angle dropped, separately from whatever happened to the maximal ceiling (Fintelman 2015) — a position that only reveals its cost on a short, hard test can still be quietly taxing a forty-minute threshold effort the whole way through.

Section 06Where riders misread this

Section 07Applying it with HELIOS

Section 08Bottom line

The fastest position on your bike is a trade-off, not a tuck. Getting lower reduces drag in a straight line, but past a point it reduces the power you can produce in a straight line too — and for a lot of riding below flat-TT speeds, the "better" aero position is measurably the slower one once you price in what it costs you physiologically.

A note on limits. This is a trade-off, not a verdict against low positions — the power penalty is partly trainable, and riders who commit to an aggressive position for weeks can claw some of it back. The right answer is individual: your flexibility, your event length, your course profile. Long, flat, fast TTs weight the aero side of the equation harder than a short hilly one does. None of this says aero doesn't matter. It says the fastest position is the one that survives contact with your own physiology, not the one with the best number on a printout.