This study was performed to determine whether Viso (the maximum speed of shortening extrapolated from force-velocity curves) equalled Vu (the unloaded speed of shortening determined by the slack test) in both living fibres from Rana temporaria and mechanically skinned fibres from Rana pipiens. In living fibres (R. temporaria) we obtained improved estimates of Viso by performing force clamps (isotonic) and length ramps (isovelocity) down to very low loads (0.005 isometric tension, P0). Force-velocity characteristics determined by force clamps and length ramps were the same. The hyperbolic Hill curves deviated from the force-velocity data at both high and low loads and underestimated Viso by varying degrees. A better estimate of Viso was obtained by linear extrapolation of data at loads from 0.005-0.02 P0 and the mean Viso at 7.5 degrees C was 4.08 muscle lengths/s +/- 0.11 (mean +/- S.E., n = 14). Improved estimates of Vu in living fibres were obtained by photographically calibrating the slack test. The mean Vu was 4.05 muscle lengths/s +/- 0.13 (mean +/- S.E., n = 14) and the intercept was 0.0156 fibre lengths (L0) +/- 0.0013 (mean +/- S.E., n = 14). The step-ramp photographic method, in which the motor speed is matched to Vu, was developed as an independent way to measure Vu in living fibres. Vu measured in this way agreed well with Vu measured by the slack test. In all living fibres, the improved estimates of Vu agreed well with the improved estimates of Viso. Vu/Viso = 0.99 +/- 0.01 (mean +/- S.E., n = 14). In mechanically skinned R. pipiens fibres, force clamps were performed down to loads of 0.01 mN. The force-velocity curve of the skinned fibres differed in shape from that of the living fibres. Although there was significant deviation from the Hill equation at low loads, the data at high loads were well fitted by the Hill curve. Viso determined by extrapolating the Hill equation to zero load was 5.87 muscle lengths/s +/- 0.38 (mean +/- S.E., n = 9) at 7.5 degrees C. In five fibres, the linear extrapolation of low loads (0.01-0.05 P0) showed that the Hill equation underestimated the true Viso by 6%. The slack test with mechanically skinned fibres was calibrated by taking a series of photographic exposures of the fibre at various times following each length step. Vu = 6.12 muscle lengths/s +/- 0.44 (mean +/- S.E., n = 10) and the intercept was 0.0585 L0 +/- 0.0069 (mean +/- S.E., n = 10).(ABSTRACT TRUNCATED AT 400 WORDS)