Background: The force–velocity relationship suggests that maximal power (Pmax) can only be produced in optimal torque (Topt) and cadence (Copt). However, the cadence at which mean maximal power (MMP) is produced has never been studied. This study aimed to determine the individual MMP–cadence relationship from in situ data. Method: We analyzed 1 year of data from 14 under-19 cyclists and calculated the MMP for each cadence between 50 and 120 rpm for 2-, 5-, and 20-minute durations. The MMP–cadence relationship was fit with a second-order polynomial function. The goodness of fit (r2) and odd-day–even-day absolute and relative reliability were evaluated, respectively, for Pmax, Topt, and Copt. Results: The goodness of fit was very high for every duration studied. Topt and Pmax, but not Copt, were significantly higher for shorter durations. Pmax was significantly correlated only with Topt for the 3 durations (r2 = .63, .71, and .64 for 2, 5, and 20 min, respectively). Discussion: Evaluation of the MMP–cadence relationship from in situ data is feasible and reliable for 2-, 5-, and 20-minute durations. This profiling approach would enable better detection of the strengths and weaknesses of cyclists and make it possible to design more effective training interventions. Practical Applications: The analysis makes it possible to identify the torque versus cadence component that individually limits power production. Knowing the Copt for a given duration of maximal effort could help athletes choose the right gear ratio and regulate cadence during a race in order to maximize performance.