temporaria (0.9-2.5degrees C) were stimulated to produce a 1 s isometric tetanus at regular intervals until constant mechanical responses were attained. Various degrees of force depression ("fatigue") were produced by decreasing the contraction interval from 30 or 15 min (control to 120, 60, 30 and 15s, respectively. In this was the steady-state tetanic force could be reversibly reduced to approximately 70% of the control value. The velocity of shortening at zero load, V0, was determined at each level of fatigue using an approach for direct measurement of V0. V0 was not significantly affected as long as the decrease in force was less than 10%. With further reduction of the isometric tension there was a progressive decline of V0 according to the following empirical relationship between percentage depression of force (delta P0) and maximum speed (delta V0) of shortening: delta V0 = 0.006 delta P02.48- 1.0 (correlation coefficient, 0.86). Cine photographic recording of nylon markers on the fibre surface provided evidence that fatigue developed uniformly along the fibre with no sign of failure of excitation in any segment. The change in mechanical performance during fatigue could be reproduced inthe non-fatigued fibre by reducing the pH of the external medium within the range 8.0-6.6 using a bicarbonate-CO2 buffer. A decrease in pH thus reduced both the rate of rise and the total amplitude of isometric force and prolonged the relaxation phase. Furthermore, there was a drop in V0 that was related to the force decline in approximately the same way as observed during fatigue. The results support the idea the fatigue involves both a reduced state of activation of the contractile system and a specific (activation independent) inhibition of crossbridge turnover. Increased intracellular H+ concentration is likely to contribute to the development of both these effects during fatigue.