Hip hemiarthroplasty (HHA) and total hip arthroplasty (THA) infections are usually considered as one entity; however, they may show important differences. We analyze these differences, as well as predictors of treatment failure (TF) and poor functional status among patients with prosthetic hip infections (PHIs). A multicenter cohort study of consecutive patients with PHIs was performed. The main outcome variable was TF after the first surgical treatment performed to treat the infection. Multivariate analysis was used to identify predictors of TF. A total of 127 patients with PHI were included (43 HHA, 84 THA). Patients with HHA infections were more frequently women (88% vs. 54%; p < 0.001), had comorbidities (86% vs. 67%, p = 0.02), and were older (median age 79 vs. 65 years, p < 0.001), and the reason for arthroplasty was more frequently a fracture (100% vs. 18%, p < 0.001). Failure of initial treatment and crude mortality were more frequent among HHA patients (44% vs. 23%, p = 0.01 and 28% vs. 7%, p = 0.001, respectively). However, HHA was not associated with TF in the multivariate analysis when hip fracture was considered; thus, variables independently associated with TF were hip fracture, inadequate surgical management, prosthesis retention, and higher C-reactive protein level. Failure of the first surgical treatment was associated with poorer functional status. HHA and THA infections showed significant differences in epidemiology, clinical features, and outcome. Although patients with HHA infections had a higher risk of TF, this was related to the reason for hip implant: a hip fracture. Success of the initial management of infection is a predictor of better clinical and functional outcome.