The early, rapid phase of tension recovery (phase 2) after a step change in sarcomere length is thought to reflect the force-generating transition of myosin bound to actin. We have measured the relation between the rate of tension redevelopment during phase 2 (r), estimated from the half-time of tension recovery during phase 2 (r = t0.5(-1)), and steady-state force at varying [Ca2+] in single fibers from rabbit psoas. Sarcomere length was monitored continuously by laser diffraction of fiber segments (length approximately 1.6 mm), and sarcomere homogeneity was maintained using periodic length release/restretch cycles at 13-15 degrees C. At lower [Ca2+] and forces, r was elevated relative to that at pCa 4.0 for both releases and stretches (between +/- 8 nm). For releases of -3.4 +/- 0.7 nm.hs-1 at pCa 6.6 (where force was 10-20% of maximum force at pCa 4.0), r was 3.3 +/- 1.0 ms-1 (mean +/- SD; N = 5), whereas the corresponding value of r at pCa 4.0 was 1.0 +/- 0.2 ms-1 for releases of -3.5 +/- 0.5 nm.hs-1 (mean +/- SD; N = 5). For stretches of 1.9 +/- 0.7 nm.hs-1, r was 1.0 +/- 0.3 ms-1 (mean +/- SD; N = 9) at pCa 6.6, whereas r was 0.4 +/- 0.1 ms-1 at pCa 4.0 for stretches of 1.9 +/- 0.5 (mean +/- SD; N = 14). Faster phase 2 transients at submaximal Ca(2+)-activation were not caused by changes in myofilament lattice spacing because 4% Dextran T-500, which minimizes lattice spacing changes, was present in all solutions. The inverse relationship between phase 2 kinetics and force obtained during steady-state activation of skinned fibers appears to be qualitatively similar to observations on intact frog skeletal fibers during the development of tetanic force. The data are consistent with models that incorporate a direct effect of [Ca2+] on phase 2 kinetics of individual cross-bridges or, alternatively, in which phase 2 kinetics depend on cooperative interactions between cross-bridges.