This research was supported by a Duke University Research Council Grant. The purpose of this study was to examine mode and speed specificity of strength training by comparing concentric and eccentric isokinetic exercise of the quadriceps. Forty-eight healthy men (mean age = 23.9 years) were randomly assigned to one of three groups: concentric training (C), eccentric training (E), or control (K). Average force (in Newtons) of 3 concentric and of 3 eccentric quadriceps contractions on the KIN-COM(R) dynamometer at 60, 120, and 180 degrees /sec was evaluated prior to and following a 6 week period during which only the C and E groups trained. Training sessions (3/week) included 4 submaximal and 1 maximal warm-up followed by 10 maximal effort isokinetic contractions of the quadriceps at 120 degrees /sec for each leg. Group C subjects trained concentrically only while Group E subjects trained eccentrically only. A t-test for independent means showed no significant right/left differences. ANOVA and Scheffe's F-tests were then used to assess the differences in training effects among the 3 groups for the left leg only. Results showed that although Group C increased slightly in both concentric and eccentric force at all speeds, the gains were significant only for concentric force at 180 degrees /sec. Group E showed significant gains (p < 0.05) in eccentric force at all speeds but not in concentric force. The K group had no significant change in concentric or eccentric force at any speed. We conclude that the eccentric mode of isokinetic exercise has highly specific strength training effects while the concentric mode has less specific training effects. In addition, speed of exercise does not appear to have specific training effects. J Orthop Sports Phys Ther 1989;11(2):70-75.