PURPOSE: This study compared the effect of an initial exercise consisting of either low-intensity eccentric or maximal isometric contractions (ISOs) on protective effect against maximal eccentric contraction (MaxECC)-induced muscle damage. METHODS: Untrained young men were placed into one of five groups (n = 13 per group): MaxECC, 10% ECC, 20% ECC, 90° ISO, and 20° ISO. The MaxECC, 10% ECC, and 20% ECC groups performed 30 ECCs of the elbow flexors using a dumbbell equivalent to 100%, 10%, and 20% of maximal voluntary isometric contraction strength, respectively. The 90° ISO and 20° ISO groups performed 30 ISOs at 90° and 20° of elbow flexion, respectively. Three weeks later, all subjects performed 30 MaxECCs with the arm used for the first bout. Changes in maximal voluntary isometric and concentric contraction strength, range of motion, upper arm circumference, plasma creatine kinase and myoglobin concentration, and muscle soreness before and for 5 d after the first and second exercise bouts were compared among groups by a two-way repeated-measure ANOVA. RESULTS: Changes in all measures after the first bout were smaller (P < 0.05) for 10% ECC, 20% ECC, 90° ISO, and 20° ISO groups compared with MaxECC group, and the changes were smaller (P < 0.05) for 10% ECC and 90° ISO than 20° ISO and 20% ECC groups. When compared with the first bout of MaxECC group, changes in the measures after the second bout were smaller for 20% ECC and 20° ISO groups with greater protective effect evident for 20° ISO group, but the protective effect conferred by these was smaller (P < 0.05) compared with MaxECCs. CONCLUSION: These results suggest that there is threshold intensity for ECCs to confer protective effect, and ISOs at a long muscle length provide preconditioning effect.