During maximal exercise, the high demand for respiratory muscle blood flow compromises blood flow to working limb locomotor muscles because of sympathetically mediated vasoconstriction (Harms et al., 1997). We hypothesised that partially unloading the inspiratory muscles would ameliorate exercise-induced quadriceps muscle fatigue and, conversely, that inspiratory muscle loading would exacerbate quadriceps fatigue. Following local ethics committee approval and written informed consent, 8 male cyclists (mean±S.E.M. peak oxygen uptake [VO_2peak] 57.8±3.5 ml kg^-1 min^-1) exercised at ≥90% VO_2peak to exhaustion (CTRL). At a separate visit, subjects exercised for the same time as during CTRL (13.2±0.9 min) but the work of breathing (WOB) was reduced using a proportional assist ventilator (PAV), which delivered positive pressure during inspiration in proportion to airflow and volume. Subjects also exercised to exhaustion (9.2±1.4 min; P<0.05 vs. CTRL) on two separate occasions while WOB was increased via inspiratory resistive loads (IRL). Two additional exercise tests were performed whereby subjects exercised for the same time as that achieved during IRL but with breathing unimpeded (IRL-CTRL). Quadriceps twitch force, in response to supramaximal paired magnetic stimuli of the femoral nerve (1-100 Hz), was assessed pre- and up to 70 min post-exercise. The flow-resistive WOB, assessed as the integrated area of the oesophageal pressure-tidal volume loop, was 56±6% lower with PAV and 79±12% higher with IRL compared with control iso-time values. The post-exercise decrease in quadriceps twitch force was attenuated with PAV (-20±5% for PAV vs. -28±5% for CTRL; P<0.05, paired-samples t-test). Repeated measures ANOVA (visit×perturbation) identified a significant within-group effect of IRL upon quadriceps twitch force (-20±7% for IRL vs. -13±8% for IRL-CTRL; P<0.05). The blood lactate response to exercise was not different (P>0.05) between either IRL and IRL-CTRL (9.0±0.7 vs. 9.2±0.7 mM) or PAV and CTRL (9.2±0.6 vs. 10.0±1.1 mM). Perceptual ratings of dyspnoea and limb discomfort (Borg CR10) were reduced with PAV and increased with IRL (P<0.05). The present results indicate the importance of WOB on locomotor muscle fatigue. The increase and decrease in quadriceps fatigue with inspiratory muscle loading and unloading, respectively, may have been due to a sympathetically mediated redistribution of blood flow in response to changes in the WOB (Harms et al., 1997). The effect of WOB on muscle fatigue may also explain our previous finding that WOB affects exercise performance (Harms et al., 2000).