Seven normal human subjects inspired a CO2-O2 mixture from a constant-flow generator while performing maximal inspiratory maneuvers from functional residual capacity. End-tidal CO2 (ETCO2) was maintained at either 5.5 (normocapnia), 3.5 (hypocapnia), or 7% (hypercapnia) on separate testing days. Subjects attained maximal mouth pressure (Pm) while breathing at either 1.25 or 1 l/s, utilizing a fixed breathing pattern (duty cycle 0.43) with an inspiratory time of 1.5 s. Maximal Pm was measured at rest and then during a 10-min endurance trial in which subjects repeated maximal voluntary inspirations with constant flow and breathing pattern. The endurance Pm data were fit to nonlinear exponential regression. The results indicated that 1) maximal Pm at rest was unaffected by changing ETCO2; 2) the rate of Pm decay over time was accelerated by hypercapnia, whereas hypocapnia showed no consistent effects; and 3) "sustainable" Pm, attained toward the end of the endurance trial, was not decreased; therefore sustainable force output was preserved in response to changing ETCO2.