This study compared physiological and perceptual responses between cycling prescribed using fixed‐power (PWR) and fixed rating of perceived exertion (RPE), when performed with blood flow restricted (BFRPWR and BFRRPE) and unrestricted (CONPWR and CONRPE). Endurance cyclists/triathletes cycled for 10 min in four separate randomized conditions; that is, two methods of prescribed exercise intensity (power at the first ventilatory threshold or RPE matched to CONPWR) combined with two occlusion levels (with BFR or without). Cardiorespiratory and perceptual variables were recorded every 2 min. Blood lactate concentration was measured pre‐, immediately and 2‐min postexercise. Power output during BFRRPE was lower than CONRPE (−13 ± 13%). The greatest physiological and perceptual responses were achieved during BFRPWR. Heart rate during BFRRPE was not different compared with CONPWR, yet was greater than CONRPE (+4 ± 11%). Muscular discomfort during BFRRPE was greater than CONPWR (+43 ± 18%) and CONRPE (+65 ± 58%). Cuff pain was greater during BFRPWR than BFRRPE (+14 ± 21%). Blood lactate concentration was not different between BFRRPE, CONPWR, and CONRPE at any timepoint. The reduction in power (fixed‐RPE trials; BFR minus unrestricted) correlated with changes in the respiratory rate (r = 0.85, confidence intervals [CI] = 0.51, 0.96) and postexercise lactate (r = 0.75, CI = 0.27, 0.93) but not muscular discomfort (r = 0.18, CI = −0.47, 0.71). Cardiorespiratory and metabolic stress, muscular discomfort, and cuff pain likely mediated self‐regulating fixed‐RPE cycling with BFR. While cycling with BFR at a fixed‐RPE resulted in less physiological stress compared to BFRPWR, it still provided a heightened level of physiological stress, with less pain and discomfort. As such, fixed‐RPE can be a suitable alternative for prescribing BFR to trained cyclists.