The purpose of this study was to determine the minimum number of contractions that are needed to detect an increase in the muscle proton spin-spin relaxation time (T2) at a given exercise intensity. Five healthy human subjects performed five sets of an exercise that included concentric and eccentric contractions of the elbow-flexor muscles with loads that were 25 or 80% of maximum. With the 80% load, the five sets involved 1, 2, 5, 10, or 20 repetitions of the exercise; with the 25% load the five sets were 2, 5, 10, 20, or 40 repetitions. The upper arm of each subject was imaged before and immediately after each set of the exercise. Spin-echo images (repetition time/echo time = 2,000 ms/30, 60, 90, and 120 ms) were collected using an extremity coil, and T2 values were calculated. The signal intensity was measured from the elbow-flexor and -extensor muscles and from the bone marrow of the humerus. With the 80% load, T2 increased in the short head of the biceps brachii after two repetitions of the elbow exercise and after five repetitions in the brachialis and the long head of the biceps brachii. With the 25% load, T2 became longer after five repetitions of the exercise for the short head of the biceps brachii and after 10 repetitions for the brachialis and the long head of the biceps brachii. T2 varied linearly with the number of contraction repetitions for each of the elbow-flexor muscles at either load (r2 > or = 0.97, P < 0.001).(ABSTRACT TRUNCATED AT 250 WORDS)