We tested the hypothesis that, in healthy middle-aged subjects ( n=11, age 51.0 +/- 3.0 years, x +/- SD), the effects of exercise training on pulmonary O(2) uptake (VO(2)) on- and off-kinetics would appear earlier than those on peak. The subjects underwent a standard training program (combined endurance and resistance training) in a health club, and were evaluated before training ("time 0", T0), and after 7 (T7), 15 (T15), 30 (T30), 60 (T60) and 90 (T90) days of training. Breath-by-breath pulmonary O(2) uptake (VO(2)), heart rate (HR), systolic (SBP) and diastolic blood pressure, and capillary blood lactate concentration ([La](b)) were determined at rest and at each workload (w during a cycle ergometer incremental exercise test. The "heart rate x blood pressure product" was calculated as (HR x SBP). The day following the incremental test, the subjects performed three repetitions of a square-wave exercise at 50% of VO(2), for the determination of pulmonary VO(2) on- and off-kinetics. VO(2) and [La](bpeak) tended to increase with training; the increases became significant at T60 or T90. HR(peak)and (HR x SBP)(peak) were unaffected by training. The time constant of the "primary" component of the VO(2) on-kinetics (tau(2)) was 46.9 +/- 17.3 s (T0), 38.1 +/- 14.2 s (T7), 34.4 +/- 12.6 s (T15), 28.8 +/- 6.8 s (T30), 30.2 +/- 8.0 s (T60), and 30.4 +/- 12.4 s (T90); a significant difference compared to T0 was observed from T15 onward. From T15 onward, tau(2) were not significantly different from values obtained (29.2 +/- 5.3 s) from a group of healthy untrained young controls ( n=7, 21.6 +/- 0.5 years). The same pattern of change as a function of training was described for the VO(2) off-kinetics. It is concluded that in 50-year-old subjects VO(2) on- and off-kinetics are more sensitive to exercise training than other physiological variables determined at peak exercise.