The purpose of this clinical study with a matched-pair design was to compare the functional short-term results obtained in patients with primary osteoarthritis of the shoulder treated with cementless surface replacement of the humeral head (CUP) with those obtained after total shoulder replacement (TSR). Twenty-two patients (average age 61.4 years; 11 men, 11 women) with primary osteoarthritis who obtained surface replacement of the humeral head were matched to a control group of 22 patients (average age 61.1 years, 11 men, 11 women) with the same in- and exclusion criteria that were treated with a total shoulder arthroplasty. Preoperative status, perioperative results and postoperative status (Constant score and subcategories, subjective status, range of motion, radiographic evaluation) were compared in all patients and controls. Non-parametric analyses were used to compare the results in both groups. Patients in the CUP group showed significantly better perioperative results (time of surgery, blood loss, days of inpatient treatment) compared to the patients in the TSR group. Results at 6 and 12 months revealed significant improvement of clinical function, significant pain reduction, and high subjective satisfaction rates in both groups. Concerning the absolute values, there were tendencially better results obtained in the TSR group compared to the CUP group at 12 months follow-up (Constant score 67.23 +/- 11.71 vs. 59.25 +/- 14.53), but only the criteria "mobility" and "abduction" revealed statistical significance. Regarding the relative improvement at 12 months compared to the baseline status, patients treated with TSR showed a significant better benefit in the total Constant score (+41.32 +/- 0.04 vs. +26.16 +/- 8.82) and in range of motion. Regarding the subjective assessment, there was no statistically significant difference in patients with TSR (mean value 1,5 (+/-0.55)) and patients with CUP (mean value 2,3 (+/-0.92)). Two CUP implants had to be removed during the follow-up period owing to secondary glenoidal erosion. At short-term follow-up, surface replacement as a technically less demanding technique provided only slightly inferior results to TSR. We therefore believe that CUP arthroplasty is a therapeutic option and can be recommended in patients with primary osteoarthritis when limited to strictly defined indications.