Every year, more than 9000 patients undergo hip fracture surgery in Norway, and about 3500 of these receive a hemiarthroplasty (HA) for a femoral neck fracture (FNF). Despite the high number of patients and extensive research, there is still no consensus on which surgical approach, fixation method, and cemented stem design to use. Several national and international guidelines on treatment options exist, but recommendations are not consistently followed. A FNF patient in Norway has an average age of 80 years and one-year mortality is reported to be 25%. Efforts should be made to optimize treatment for this high-risk patient group. Based on data from the Norwegian Hip Fracture Register (NHFR), we have investigated whether surgical approach, method of stem fixation or type of femoral stem influenced the risk of reoperation, mortality, and patient-reported outcome measures (PROMs) in patients treated with HA. In Paper I, we included patients aged 60 years and older with FNF treated from 2005 to 2014. In all, 18,918 HA procedures were reported with direct lateral approach (DLA) and 1,990 with posterior approach (PA). There were statistically significant differences in PROMs with less pain, better satisfaction, better quality of life and fewer patients having walking problems after surgery with PA than with DLA. However, using a Cox regression model adjusted for confounding variables, we found no difference in risk of reoperation between DLA and PA (HRR 1.2; 95% CI 0.9-1.4; p = 0.2) with DLA as reference. In Paper II, a total of 7,539 uncemented HAs and 22,639 cemented HAs for FNF in patients 70 years or older treated in 2005-2017 were compared for risk of reoperation, mortality rate, and PROMs. Uncemented HAs had a higher overall risk of reoperation for any reason (HRR 1.5; 95% CI 1.4-1.7; p < 0.001). Although higher early mortality was found for those receiving cemented implants, no differences were found in the overall one-year mortality rate (HRR 1.0; 95% CI 0.9-1.0; p = 0.12). HA fixation type was not associated with differences in patients’ pain (19 versus 20 for uncemented and cemented HAs respectively, p = 0.052) or quality of life (EQ-VAS score 64 versus 64, p = 0.43, EQ-5D index score 0.64 versus 0.63, p = 0.061), one year after surgery. In Paper III, the different types of cemented stems were studied. A total of 20,529 primary cemented hemiarthroplasties for FNF in patients aged 70 years or older treated in 2005-2016 were included. Polished tapered stems (n=12,064) (the Exeter and CPT prostheses), straight stems (n=5,543) (the Charnley, Charnley Modular, and Spectron EF prostheses), and anatomic stems (n=2,922) (the Lubinus SP2 prosthesis) were compared. When dividing the stems according to design, better survival for the stems with a straight design (HRR 0.66; 95% CI 0.55 to 0.79; p < 0.001) and with an anatomic design (HRR 0.74; 95% CI 0.59 to 0.93; p = 0.010) was found compared to the polished tapered stem design. Reoperation due to periprosthetic fracture (PPF) occurred almost exclusively after surgery with polished tapered stems. In conclusion, patients operated for FNF with HA performed with a PA reported less pain, better patient satisfaction, better quality of life and fewer walking problems compared to DLA. No differences in risk of reoperation between the surgical approaches were found. Uncemented HAs had a greater reoperation risk than cemented. The fixation method did not influence pain, quality of life, or the one-year mortality rate after HA. In cemented HAs, differences in reoperation rates seemed to favour anatomic and straight stems over polished tapered stems, which had a higher risk of PPF.