Introduction This study explored the effects of general (GA) and neuraxial (NA) anesthesia on the outcomes of primary total joint replacement (TJR) in terms of postoperative mortality, length of stay (LOS), and hospital treatment costs. Methods From 1997 to 2010, this nationwide population-based study retrospectively evaluated 7,977 patients in Taiwan who underwent primary total hip or knee replacement. We generated two propensity-scorematched subgroups, each containing an equal number of patients who underwent TJR with either GA or NA. Results Of the 7,977 patients, 2,990 (37.5%) underwent GA and 4,987 (62.5%) underwent NA. Propensity-score matching was used to create comparable GA and NA groups adjusted for age, sex, comorbidities, surgery type, hospital volume, and surgeon volume. Survival over the first three years following surgery was similar. The proportion of patients alive up to 14 years postoperatively for those undergoing NA was 58.2% (95% confidence interval [CI] 50.4 to 66.0), and for those undergoing GA it was 57.3% (95% CI 51.4 to 63.2). Neuraxial anesthesia was associated with lower median [interquartile range; IQR] hospital treatment cost ($4,079 [3,805-4,444] vs $4,113 [3,812-4,568]; P \ 0.001) and shorter median days vs 8 [6-10] days, respectively; P = 0.024). Conclusions Our results support the use of NA for primary TJR. The improvements in hospital costs persist even when anesthesia costs are removed. The mechanism underlying the association between NA and long-term survival is unknown.