Radial head fractures are the most frequent fractures of the elbow joint in adults. For Mason typeII fractures without concomitant injuries favorable results have been shown with operative and conservative management. There is insufficient evidence concerning elbow joint stability after conservative treatment compared to open reduction and internal fixation (ORIF). All patients with isolated Mason type II (two part fracture displaced >2 mm and <5 mm) radial head fractures between 1 January 2003 and 1 April 2013 were retrospectively reviewed. Exclusion criteria were age <18years, associated fractures of the ipsilateral extremity or elbow luxation. A total of 50patients (mean age 44.2years, range 19-71years) who received either ORIF (n = 31) or conservative treatment (n = 19) were included. The mean follow-up was 43.2months (range9-61 months). Patients were evaluated using the Disability of the Arm, Shoulder and Hand (DASH) score, Mayo Elbow Performance score (MEPS), and Oxford Elbow score (OES). Joint stability (varus and valgus stress) was assessed under fluoroscopy and a distinction was made between slight instability (angulation <10°), moderate instability (angulation ≥10°) and gross instability (elbow dislocation). Residual fracture displacement (conservative: 2.7 mm, ORIF: 1.4 mm, p < 0.042) and varus/valgus joint stability (3% ORIF vs. 26% conservative, p = 0.031) showed significant differences. The stability as tested by a radiological dynamic procedure showed an instability after ORIF in 3% of the joints compared to 26% after conservative treatment; however, this did not influence the short to mid-term clinical outcome: No significant differences were found in the DASH score (conservative 33points, ORIF 36points), MEPS (conservative 76points, ORIF 78points) and OES (conservative 41points, ORIF 43points). Both conservative management and operative treatment had agood functional outcome. Operative treatment showed apositive tendency concerning radiological and functional outcome without statistical significance.