In a randomized cross-over study of 15 healthy men aged 20–30 years, we compared physiological and perceptual responses during treadmill and cycle exercise test protocols matched for increments in work rate — the source of increased locomotor muscle metabolic and contractile demands. The rates of O2 consumption and CO2 production were higher at the peak of treadmill versus cycle testing (p ≤ 0.05). Nevertheless, work rate, minute ventilation, tidal volume (VT), breathing frequency (fR), inspiratory capacity (IC), inspiratory reserve volume (IRV), tidal esophageal (Pes,tidal) and transdiaphragmatic pressure swings (Pdi,tidal), peak expiratory gastric pressures (Pga,peak), the root mean square of the diaphragm electromyogram (EMGdi,rms) expressed as a percentage of maximum EMGdi,rms (EMGdi,rms%max), and dyspnea ratings were similar at the peak of treadmill versus cycle testing (p > 0.05). Ratings of leg discomfort were higher at the peak of cycle versus treadmill exercise (p ≤ 0.05), even though peak O2 consumption was lower during cycling. Oxygen consumption, CO2 production, minute ventilation, fR, Pes,tidal, Pdi,tidal and Pga,peak were higher (p ≤ 0.05), while VT, IC, IRV, EMGdi,rms%max, and ratings of dyspnea and leg discomfort were similar (p > 0.05) at all or most submaximal work rates during treadmill versus cycle exercise. Our findings highlight important differences (and similarities) in physiological and perceptual responses at maximal and submaximal work rates during incremental treadmill and cycle exercise testing protocols. The lack of effect of exercise test modality on peak work rate advocates for the use of this readily available parameter to optimize training intensity determination, regardless of exercise training mode.