Purpose Comparing the MRI features of the grafts between a group of patients treated with an over-the-top anterior cruciate ligament reconstruction technique that preserves the hamstring attachment and a control group with a classical reconstruction technique. Methods Patients were assigned to a standard reconstruction technique or an Over-the-top plus lateral plasty technique. All patients underwent preoperative, 4-months and 18-months MRI; together with a clinical evaluation with KOOS and KT1000 laxity assessment. MRI study involved different parameters: the “Graft” was evaluated with the continuity, Howell Grading system, presence of liquid and signal noise quotient. The “Tibial Tunnel” was evaluated with the signal noise quotient, presence of edema or liquid and tunnel widening. All points assigned to each parameter formed a composite score ranging from 0–10. Tunnel and graft positioning were evaluated. Results At 18-month 20 MRIs (10 each group) were available, demographics were not significantly different between groups. The non-detached group showed significantly less liquid within the graft at 4-months ( p  = 0.008) and 18-months ( p  = 0.028), the tunnel was significantly smaller ( p  < 0.05) and less enlarged at both follow-ups ( p  < 0.05), signal noise quotient of the intra-tunnel graft was lower at 18-months ( p  < 0.05). The total score of the non-detached group saw a significant improvement at 4-months ( p  = 0.006) that remained stable at 18-months (n.s.). Conclusions Hamstring grafts, which tibial insertions were preserved, showed better MRI features at 4-and 18-months follow-up, especially in terms of liquid effusion, tunnel enlargement and signal noise quotient. Level of evidence IV.