Background Stoma closure has been associated with a high rate of surgical site infection (SSI) and the ideal stoma-site skin closure technique is still debated. The aim of this study was to compare the rate of SSI following primary skin closure (PC) versus a skin-approximating, subcuticular purse-string closure (APS). Methods All consecutive patients undergoing stoma closure between 2002 and 2007 by two surgeons at a single tertiarycare institution were retrospectively assessed. Patients who had a new stoma created at the same site or those without wound closure were excluded. The end point was SSI, determined according to current CDCguidelines, at the stoma closure site and/or the midline laparotomy incision. Results There were 61 patients in the PC group (surgeon A: 58 of 61) and 17 in the APS group (surgeon B: 16 of 17). The two groups were similar in baseline and intraoperative characteristics, except that patients in the PC group were more often diagnosed with benign disease (p = 0.0156) and more often had a stapled anastomosis (p = 0.002). The overall SSI rate was 14 of 78 (18%). All SSIs occurred in the PC group (14 of 61 vs. 0 of 17, p = 0.03). Conclusions Our study suggests that a skin-approximating closure with a subcuticular purse-string of the stoma site leads to less SSI than a primary closure. Randomized studies are needed to confirm our findings and assess additional end points such as healing time, cost, and patient satisfaction.