Slowing of relaxation is one of the anticipated changes in the contraction of fatigued skeletal muscle. However, interpretation of the mechanism(s) contributing to slowed relaxation may be affected by the measurement technique employed. In this study, relaxation was measured in three ways: (i) traditional half-relaxation time; (ii) peak rate of relaxation; and (iii) late relaxation time, measured from 50 to 25% of peak developed tension. When rat gastrocnemius muscle was stimulated indirectly in situ at 10 Hz, developed tension increased in 10 s to 185%, then decreased to 39% after 1 min with little additional change over the next 4 min. After 10 s of inactivity, developed tension was 60% of the initial value, but did not recover further over the next 20 min. The half-relaxation time transiently decreased at the start of stimulation, then by 20 s was considerably prolonged. Within 10 s of recovery, half-relaxation time returned to prestimulation values but became prolonged again by 10 min of recovery. The peak rate of relaxation was proportional to the developed tension at all times except 2.5-10 s of 10 Hz stimulation, at which time acceleration of relaxation was evident, and 15-20 s of the 10 Hz stimulation when it was relatively decreased. The late relaxation time increased during the repetitive stimulation, returned to control level early in recovery, then increased again, by 5 min of recovery. The diverse responses indicated by these indices of relaxation potentially discriminate different mechanisms which contribute to slowing of relaxation in fatigue, a point which would be missed if a single method of measurement of relaxation was employed.