The influence of stimulation frequency on efficiency (= total work output/high-energy phosphate consumption) was studied using in situ medial gastrocnemius muscle tendon complexes of the rat. The muscles performed 20 repeated concentric contractions (2/s) at 34 degrees C. During these repeated contractions, the muscle was stimulated via the severed sciatic nerve with either 60, 90, or 150 Hz. The muscle was freeze-clamped immediately after these contractions, and high-energy phosphate consumption was determined by measuring intramuscular chemical change relative to control muscles. The average values (+/-SD) of efficiency calculated for 60, 90, and 150 Hz were 18.5 +/- 1.5 (n = 7), 18.6 +/- 1.5 (n = 9), and 14.7 +/- 1.3 mJ/micromol phosphate (n = 9). The results indicate that the efficiency of the muscles that were submaximally activated (60 or 90 Hz) was higher (+26%, P < 0.05) than that of those maximally activated (150 Hz). Additional experiments showed that the low efficiency at maximal activation levels is unlikely to be the result of a higher energy turnover by the Ca2+ -ATPase relative to the total energy turnover. Therefore, alternative explanations are discussed.