PurposeGraft diameter ≥ 8 mm reduces the risk of failure after anterior cruciate ligament reconstruction (ALCR) with hamstring tendon autograft. Pre‐operative measurement of gracilis (GT) and semitendinosus (ST) cross‐sectional area using MRI has been utilized but the optimal location for measurement is unknown. The main purpose of this study was to examine the cross‐sectional areas of GT + ST at different locations and develop a model to predict whether a doubled hamstring graft of GT + ST will be of sufficient cross‐sectional area for ACLR. MethodsA retrospective review was performed of 154 patients who underwent primary ACLR using doubled hamstring autograft. Cross‐sectional area measurements of GT + ST on pre‐operative MRI axial images were made at three locations: medial epicondyle (ME), tibiofemoral joint line (TJL), and tibial physeal scar (TPS) and calculated the correlation of intra‐operative graft size for each location using the Pearson’s correlation coefficient. A receiver operating characteristic (ROC) established a threshold that would predict graft diameter ≥ 8 mm. ResultsMeasurement of GT + ST at the ME had a stronger correlation (r = 0.389) to intra‐operative graft diameter than measurements at the TJL (r = 0.256) or TPS (r = 0.240). The ROC indicated good predictive value for hamstring graft diameter ≥ 8 mm based on MRI measurement at the ME with the optimal threshold with the highest sensitivity and specificity as 18 mm2. ConclusionCross‐sectional area measurement of GT + ST at the ME correlated most closely to intra‐operative diameter of a doubled hamstring autograft compared to measurements at the TJL or the TPS. As graft diameter < 8 mm is correlated with higher failure rates of ACL surgery, the ability to pre‐operatively predict graft diameter is clinically useful. Level of evidenceLevel III, prognostic study.