Surgical learning curves are a representation of the number of cases required for a surgeon to reach a stable rate of outcomes and complications. In this study, the authors present the learning curve for the direct superior (DS) approach to total hip arthroplasty, which is a muscle-sparing modification to the mini-posterior technique. This was a retrospective analysis of the first 40 primary DS cases done by a single surgeon. These cases were divided into 2 groups of 20 and compared for intra- and postoperative complications, acetabular component positioning, and Harris Hip Score at 90 days after surgery. As a control, the first 20 primary mini-posterior cases were analyzed as the baseline performance of the surgeon and the DS approach. There was no statistically significant difference between the first and second sets of DS patients regarding Harris Hip Score, intraoperative complications, dislocations, estimated blood loss, length of hospital stay, and components positioned within the Lewinnek safe zone. The only statistically significant difference between the first and second sets of DS cases was a decrease in operating time (P<.001). This suggests that the learning curve is less than 20 patients. The results also indicate that the first 20 DS cases ambulated farther (P=.007) and had a shorter length of stay (P=.007), outperforming the mini-posterior approach. This study suggests that the DS learning curve is short for surgeons trained in the posterior approach. The data could be especially pertinent as 90-day results and safety are becoming increasingly important in evaluating performance for bundled-payment models. [Orthopedics. 2020;43(x):xx-xx.].Copyright 2020, SLACK Incorporated.