On transition to moderate intensity exercise (i.e. below the lactate threshold; LT) the rate of adaptation (τ) of pulmonary O₂ uptake (VO₂) is slower when initiated from a higher compared to a lower work rate [1-3]. This has been ascribed to a slowed adaptation in skeletal muscle blood flow [2,3] or the intrinsic properties of the recruited muscles [1]. However, the role of the pre-transition metabolic rate remains unclear. That is, are the slower VO₂ kinetics a consequence of the raised pre-transition VO₂, or due to the de-novo recruitment of motor units with inherently slower activation of oxidative metabolism? To elucidate this, we determined VO₂ kinetics at the onset of a step change in work rate with and without a raised metabolic rate. We hypothesised that VO₂ kinetics would be slower in the upper compared to the lower reaches of the moderate-intensity domain [1,3] due to the intrinsically slower kinetics of the newly recruited motor units, such that the τ of the fundamental VO₂ response (τVO₂) would be insensitive to an alteration of pre-transition metabolic rate. Following informed consent, 6 healthy men completed ramp-incremental cycle ergometry to the limit of tolerance to estimate LT. Subjects then completed repeats of: 1) two equal step transitions, 20W-45%LT (S1) immediately followed by 45%-90%LT (S2); 2) two exercise bouts (R1, R2), each 20W-90%LT separated by 30s at 20W to allow VO₂ to recover to ~45%LT i.e. a pre-transition level similar to S2; and 3) two bouts of 20W-90%LT exercise separated by 12 min at 20W (F1, F2). τVO₂ was estimated from breath-by-breath measurements (mass-spectrometry and turbinometry) using non-linear least squares analysis [4], and compared by repeated-measures ANOVA. The τVO₂ was 27 ± 4 s on transition from 20W-90%LT with a raised metabolic rate (R2), which was not different (p>0.05) from τVO₂ measured during S2 (25 ± 5 s); the pre-transition VO₂ being well matched (R2: 1.31 ± 0.11 L.min^-1; S2 1.24 ± 0.06 L.min^-1; p>0.05). These were both greater (p<0.05) than τVO₂ manifest during S1 (18 ± 5 s). As expected, τVO₂ following ‘full’ recovery (F2: 24 ± 4 s) was not different from control (F1: 23 ± 7 s; R1: 23 ± 3 s). These data are consistent with previous findings [1-3] that VO₂ kinetics are slower in the upper reaches of the moderate-intensity domain, but suggest that this is consequent to the high pre-transition metabolic rate. This may be due to: 1) an increased ATP production requirement when exercise is initiated from a less favourable ΔG_ATP; or 2) a reduction in the sensitivity of ADP-stimulated VO₂ above the Km. It is assumed that the same motor units are recruited in R1 and R2 suggesting that these mechanisms operate within the active musculature. We therefore conclude that VO₂ kinetics are sensitive to the pre-transition metabolic rate from which they are estimated.