The purpose of this work was to investigate the effect of multidirectional soccer-specific fatigue on hamstring muscle strength and angle of peak torque. Sixteen male semi-professional soccer players (mean ± S.D.: age: 21.3 ± 2.9 years; height 185.0 ± 8.7 cm; body mass 81.6 ± 6.7 kg) completed the SAFT90, a multidirectional, intermittent 90-min exercise protocol based on data from English Championship soccer matches. Prior to exercise (t0), at half-time (t45) and post-exercise (t105), subjects performed three maximal dominant limb isokinetic contractions (Biodex, System 3) at 120° s−1 through a 90° range for concentric and eccentric knee flexors and concentric knee extensors. Analysis of variance revealed significant time dependant reductions in gravity corrected eccentric hamstring peak torque, and consequently in the functional hamstring:quadriceps ratio (P < 0.01). Eccentric hamstring peak torque decreased significantly during each half (t0: 272.0 ± 43.2; t45: 240.4 ± 43.3; t105: 226.3 ± 45.7 N m). The functional hamstring:quadriceps ratio also decreased significantly during each half (t0: 116.6 ± 21.2; t45: 107.1 ± 17.6; t105: 98.8 ± 20.3%). There were no significant changes in concentric hamstring or quadriceps peak torque observed during SAFT90 (P > 0.05). Data analysis also revealed significant differences for Angle of Peak Torque for eccentric hamstrings (P < 0.05) which was significantly higher at the end of each half (t45: 37 ± 15; t105: 38 ± 18°) than the pre-exercise value (t0: 28 ± 12°). There was a time dependant decrease in peak eccentric hamstring torque and in the functional strength ratio which may have implications for the increased predisposition to hamstring strain injury during the latter stages of match-play.