INTRODUCTION: We tested the linear critical power ([Formula: see text]) model for discrete incremental ramp exercise implying recovery intervals at the end of each step. METHODS: Seven subjects performed incremental (power increment 25 W) stepwise ramps to subject's exhaustion, with recovery intervals at the end of each step. Ramps' slopes (S) were 0.83, 0.42, 0.28, 0.21, and 0.08 W s-1; recovery durations (t r) were 0 (continuous stepwise ramps), 60, and 180 s (discontinuous stepwise ramps). We determined the energy store component (W'), the peak power ([Formula: see text]), and [Formula: see text]. RESULTS: When t r = 0 s, [Formula: see text] and W' were 187 ± 26 W and 14.5 ± 5.8 kJ, respectively. When t r = 60 or 180 s, the model for ramp exercise provided inconsistent [Formula: see text] values. A more general model, implying a quadratic [Formula: see text] versus [Formula: see text] relationship, was developed. This model yielded, for t r = 60 s, [Formula: see text] = 189 ± 48 W and W' = 18.6 ± 17.8 kJ, and for t r = 180 s, [Formula: see text] = 190 ± 34 W, and W' = 16.4 ± 16.7 kJ. These [Formula: see text] and W' did not differ from the corresponding values for t r = 0 s. Nevertheless, the overall amount of energy sustaining work above [Formula: see text], due to energy store reconstitution during recovery intervals, was higher the longer t r, whence higher [Formula: see text] values. CONCLUSIONS: The linear [Formula: see text] model for ramp exercise represents a particular case (for t r = 0 s) of a more general model, accounting for energy resynthesis following oxygen deficit payment during recovery.