The length dependence of force development of mammalian skeletal muscles was evaluated during twitch, double-pulse, and tetanic contractions, and the relation between muscle length and the time-dependent characteristics of twitch and double-pulse contractions were determined. In situ isometric contractions of the rat gastrocnemius muscle were analyzed at seven different lengths, based on a reference length at which the maximal response to double-pulse contractions occurred (Lopt-2P). Twitch and double-pulse contractions were analyzed for developed tension (DT), contraction time (tC), average rate of force development (DT-tC(-1)), half-relaxation time (t50%R), peak rate of relaxation (DT x dtmin(-1)), and 90%-relaxation time (t90%R). Considering the length at which maximal tetanic DT occurred to be the optimal length (Lo-TET), the peak DT for twitch contractions and double-pulse contractions was observed at Lo-TET + 0.75 mm (p < 0.05) and Lo-TET + 0.1 mm (p > 0.05), respectively. When measured at the length for which maximal twitch and double-pulse contractions were obtained, tetanic DT was 95.2 +/- 3 and 99.0 +/- 2% of the maximal value, respectively. These observations suggest that double-pulse contractions are more suitable for setting length for experimental studies than twitch contractions. Twitch and double-pulse contraction tC were 15.53 +/- 1.14 and 25.0 +/- 0.6 ms, respectively, at Lopt-2P, and increased above Lopt-2P and decreased below Lopt-2P. Twitch t50%R was 12.18 +/- 0.90 ms at Lopt-2P, and increased above Lopt-2P and below Lopt-2P. Corresponding changes for double-pulse contractions were greater. Stretching the muscle leads to slower twitch contractions and double-pulse contractions, but the mechanisms of this change in time course remain unclear.