Introduction: This study aims to investigate whether implant positioning is associated to risk of reoperation following internal fixation of undisplaced femoral neck fractures (FNF) with a posterior tilt < 20° in patients > 65 years. Method: Patients were retrieved from the Danish Multidisciplinary Hip Fracture Register in the period 2009–2013. The patients’ health records and x-rays were reviewed for age, sex, implant, Charlson Comorbidity Index, mortality, reoperation, fracture classification and implant positioning. X-rays were measured for implant positioning using a pre-existing scoring system (the Schep score) as well as some additional measurements. Primary outcome was reoperation within 2 years. The study included 406 patients, 75% females and the median (range) age was 82 (65–99) years. Odds ratios (OR) are shown with 95% confidence interval. Results: There were 45 (11%) reoperations. Six measurements were individually associated to risk of reoperation: 1) Distance to inferior calcar < 1 or ≥ 6 mm, OR 2 (1.1-4), 2) Distance to superior cortex 10 mm, OR 2(1.1-5), 3) Tip-head distance < 3 or ≥ 20 mm, OR 2 (1.1-4), 4) Placement in the superior or inferior 15 mm, OR 2 (1.1-5), 5) Placement in the anterior 25 %, OR 6 (1.8-20), 6) Inter-implant angle ≥ 5°, OR 3 (1.4-8). The Schep score had no associated to reoperation, and therefore a new implant positioning (IMPO) score was developed. The IMPO score consists of 6 items; 1 point given for each acceptable implant placement. An IMPO score less than 5 had an increased risk of reoperation; 0-2 points OR 22 (7-71) and OR 5 (2-11) for 3-4 points, compared to a score of 5-6. Among the 207 patients with a score of 5-6, the reoperation frequency was 4%. Conclusions: This study identified implant positioning as a predictor to an increased risk of reoperation in undisplaced FNF. The newly developed IMPO score seems promising for identifying risk of reoperation.