OBJECTIVE: To describe 3.0-Tesla (T) magnetic resonance imaging (MRI) findings of hamstring muscles in clinically recovered athletes who were cleared for return to play (RTP). DESIGN: Prospective observational study. SETTING: Hospital. PARTICIPANTS: Athletes (amateur and professional) who were cleared for RTP after hamstring injury. Fifty-eight participants were included in the analysis. INDEPENDENT VARIABLES: 3-T MRI at baseline (within 7 days from initial injury) and MRI at RTP (within 10 days of RTP). MAIN OUTCOME MEASURES: Injury location, grade of injury (modified Peetrons and British Athletics Muscle Injury Classification/BAMIC), presence and the extent of intramuscular signal abnormality, intramuscular tendon disruption, and thickness. Reinjuries within 1 year of RTP were recorded. RESULTS: Magnetic resonance images at RTP showed that 55 (95%) participants had intramuscular increased signal intensity (edema) and 44 (76%) participants had intramuscular abnormal low-signal intensity (suggesting fibrosis) on MRI. There was an overall reduction of injury grades according to the modified Peetrons and BAMIC classification at initial injury to RTP. Three (5%) participants had no abnormal signal intensities (grade 0 or grade 0A) on MRI at RTP. Intramuscular tendon disruption, waviness, and tendon thickening were present at RTP in, respectively, 22 (38%), 15 (26%), and 36 (62%). We recorded 3 (5%) reinjuries. CONCLUSIONS: At RTP, 3.0-T MRI shows high percentages of MRI abnormalities (edema, fibrosis, and intramuscular tendon disruption and thickening). We conclude that complete normalization of 3.0-T MRI is not expected for RTP decision after a hamstring injury. Its possible association with reinjury risk has to be determined in larger cohorts.