1. The purpose of this study was to determine whether ATP-sensitive K+ (KATP) channels are activated and contribute to the decrease in force during fatigue development in the sartorius muscle of the frog, Rana pipiens. Tetanic force (elicited by field stimulation), action potential and membrane conductance (using conventional microelectrodes), were measured in the presence and absence of glibenclamide, a KATP channel antagonist. Experiments were performed in bicarbonate-buffered solutions at pH 7-2.2. In unfatigued muscle 100 ,umol I-' glibenclamide had no effect on the resting potential, the overshoot, the half-depolarization time or the maximum rate of depolarization of action potentials, while the mean half-repolarization time increased by 19 ± 4 % (± S.E.M.) and the maximum rate of repolarization decreased by 17 + 5 %. 3. Fatigue was elicited using 100 ms tetanic contractions every I s for 3 min. In the absence of glibenclamide the mean half-repolarization time increased from 057 + 005 to 089 + 0 05 ms during fatigue. The mean half-repolarization times after fatigue, when muscle fibres were exposed to 100 ,umol I1 glibenclamide either 60 min prior to fatigue or 60 s before the end of fatigue, were 1-16 + 0-08 and 1t17 + 0 07 ms respectively.Application of 100 #smol I-' glibenclamide after 5 min of recovery did not increase the half-repolarization time, but decreased the rate of recovery compared to control values.4. In unfatigued muscles, 100 jumol I' glibenclamide did not affect the tetanic contraction. In the absence of glibenclamide, the mean tetanic force after fatigue was 11 0 + 0 9 % of prefatigue values. Application of 100 umol I' glibenclamide 60 min before fatigue increased the rate of fatigue development as the mean tetanic force was 4-8 + 0-8 % after 3 min of stimulation. The addition of 100 ,umol I-' glibenclamide 60 s before the end of fatigue had no effect on tetanic force during this time compared to control. 5. In the absence of glibenclamide, muscles recovered 90-1 + 1-6 % of their tetanic force after 100 min. Addition of 100,umol I' glibenclamide 60 min prior to fatigue significantly reduced the capacity of muscles to recover their tetanic force: after 100 min of recovery tetanic force was only 47-3 + 9.4 % of the pre-fatigue value. Application of 100 jumol I' glibenclamide 60 s prior to the end of fatigue had a much smaller effect on the recovery as 79.4 + 6'2 % of the tetanic force was recovered in 100 min. Addition of glibenclamide after 5 min of recovery had no effect. 6. The results from this study support the proposal that KATP channels are activated during fatigue and they contribute to the repolarization phase of the action potential. Although no evidence was found that activation of KATP channels during fatigue contributes to the force decrease during fatigue development, the impairment of force recovery following fatigue in the presence of glibenclamide supports the notion that KATP channels play an important protective role.