PURPOSE We have previously reported that cycling muscle recruitment is less skilled in triathletes than cyclists with similar cycling training loads. This study investigated the relationship between muscle recruitment and cadence in triathletes, which may provide further evidence of less skilled movement control. METHODS: Participants were eight elite triathletes. Comparisons were made with data from our earlier study of nine elite and ten novice cyclists. Electromyographic (EMG) recordings were made using indwelling electrodes from tibialis anterior (TA), tibialis posterior (TP), peroneus longus (PL), gastrocnemius lateralis (GL) and soleus (SOL). Cycling at individual preferred cadence (PC), 55–60, 75–80 and 90–95 rpm were investigated in random order. Cycling trials were conducted on each participant's bike secured on a magnetic trainer. Cycling intensity was controlled using ratings of perceived exertion (RPE). RESULTS The increase in cadence from 57.5 to 92.5 rpm was associated with a decrease in modulation of muscle activity (i.e. longer bursts of activity and a higher level of activity between main bursts) in triathletes for TA, SOL (p 0.56). Coactivation increased with cadence in triathletes in proportion to increased duration of primary muscle activity, as it did in novice cyclists. Peak EMG amplitude increased with cadence in triathletes and was not minimized at PC (p = 0.21), which is consistent with our findings in novice and elite cyclists. Times of peak EMG activity and variability of muscle recruitment were not influenced by cadence in any group (p > 0.32). CONCLUSION PC is not related to peak EMG amplitude in triathletes, which is consistent with our findings in novice and elite cyclists. Less modulation of muscle activity and greater muscle coactivation with increased in cadence in triathletes is considered as further evidence of less skilled control of the movement relative to cyclists with similar cycling training loads.