The aim of this study was to assess the effect of the recovery duration in intermittent training drills on metabolism and coordination in sport games. Ten nationally ranked male tennis players (age 25.3+/-3.7 years, height 1.83+/-0.8 m, body mass 77.8+/-7.7 kg; mean +/- sx) participated in a passing-shot drill (baseline sprint with subsequent passing shot) that aimed to improve both starting speed and stroke quality (speed and precision). Time pressure for stroke preparation was individually adjusted by a ball-machine and corresponded to 80% of maximum running speed. In two trials (T10, T15) separated by 2 weeks, the players completed 30 strokes and sprints subdivided into 6 x 5 repetitions with a 1 min rest between series. The rest between each stroke-and-sprint lasted either 10 s (T10) or 15 s (T15). The sequence of both conditions was randomized between participants. Post-exercise blood lactate concentration was significantly elevated in T10 (9.04+/-3.06 vs 5.01+/-1.35 mmol x l(-1), P < 0.01). Running time for stroke preparation (1.405+/-0.044 vs 1.376+/-0.045 s, P < 0.05) and stroke speed (106+/-12 vs 114+/-8 km x h(-1), P < 0.05) were significantly decreased in T10, while stroke precision - that is, more target hits (P < 0.1) and fewer errors (P < 0.05) - tended to be higher. We conclude that running speed and stroke quality during intermittent tennis drills are highly dependent on the duration of recovery time. Optimization of training efficacy in sport games (e.g. combined improvement of conditional and technical skills) requires skilful fine-tuning of monitoring guidelines.