Purpose: To report short-tem clinical results and complications in treating single compartmental knee arthrosis using unicompartmental knee arthroplasty (UKA). Methods: From 2003 to 2005, 38 patients (5 males and 33 females) with 48 knees had UKA performed. Medical records were retrospectively reviewed for body weight, body mass index (BMI), the Hospital for Special Surgery knee score (KSS), patient age, and degree of alignment change. Radiographic imaging was performed pre-and postoperatively at 1 month, 3 months and annually and these images also reviewed. Patients with periprosthetic fractures were compared to those without. Results: Seven patients had perioperative periprosthetic fracture due to the following factors: pinhole related stress (1); improper tibia osteotomy (1); osteoporotic bone stock with technical error (3); and trauma (2). Of 31 patients with more than 2-year follow-up, 28 had well-functioning prostheses along with functional improvement including pain and KSS scores. There were no statistically significant differences between the fracture and non-fracture group for age, BMI or alignment change. Three patients required revision arthroplasties, total knee arthroplasty (2) and tibia component revision (1). Review of the radiographic images showed asymptomatic nonprogressive radiolucent lines (4) and polyethylene insert wear without loosening (3). Conclusions: UKA may play a role in treating single compartment knee arthrosis; however, patient selection and surgical technique are important factors.