Background: No consensus exists about whether a volar approach (VA) or dorsal approach (DA) for proximal interphalangeal (PIP) arthroplasty yields better results. Previously reported range of motion (ROM) and complications vary from study to study. This retrospective review compared the ROM and complication rates of VA and DA approaches to PIP arthroplasty. Methods: The study included 66 adults (88 digits) who underwent PIP arthroplasty from 2000 to 2015, with minimum 30-day follow-up. Demographic data, surgical approach, pre- and post-operative ROM, duration of immobilization, timing and duration of hand therapy (occupational therapy [OT]), and major and minor complications were recorded. We compared mean change in ROM, postoperative ROM, and complication rates, and examined the association of duration of immobilization and time to OT initiation with postoperative ROM. Results: While there was no difference in postoperative ROM between volar and dorsal groups (56° and 54°, respectively, P > .05), there was a greater gain in ROM in the DA group (25° vs 2.7°, P = .017). There was no statistically significant difference in overall incidence of complications (VA: 37.8%, DA: 30.3%; P > .05) or revision surgery (VA: 15.6%, DA: 17.1%; P > .05). There were no differences in duration of immobilization, time to OT initiation, or number of OT sessions between the two groups, and none of these correlated with postoperative ROM. Conclusions: We identified no statistical difference in mean postoperative ROM, incidence of complications or revision surgery between volar and dorsal approaches for PIP arthroplasty.