Nosocomial candidemia extends the length of hospital stay, increases the costs of medical care, and is associated with a high mortality rate. Epidemiological data that assist in the choice of initial therapy may help to improve the prognosis. The present study was undertaken to investigate the epidemiology of nosocomial candidemia and identify risk factors for nosocomial candidemia caused by C. albicans and non-albicans Candida species (NAC). A retrospective chart review was undertaken to analyze cases of nosocomial candidemia treated at the Beijing Friendship Hospital between January 2008 and December 2012. All cases of candidemia were identified using the previously published criteria. Among 106 patients analyzed, 53.8% had nosocomial candidemia caused by NAC. Candida albicans was the most common causative agent, accounting for 46.2% of all cases, followed by C. glabrata (25.5%), C. tropicalis (15.1%), C. parapsilosis (10.4%) and C. Krusei (0.9%). Comparison of nosocomial C. albicans and NAC candidemia by multivariate logistic regression showed that factors independently associated with nosocomial NAC candidemia included exposure to azole agents (odds ratio [OR]: 3.359; 95% confidence interval [CI]: 1.136-10.154; P = .031) and artificial surgical implants (OR: 37.519; 95% CI: 2.5-562.998; P = .009). A significant risk factor for nosocomial C. albicans candidemia was cancer surgery (OR: 0.075; 95% CI: 0.013-0.437; P = .004). Clinical and epidemiological differences in the risk factors between nosocomial candidemia caused by C. albicans and NAC should be considered when selecting an initial antifungal regimen for the treatment of adult patients. This should be undertaken before the availability of species identification and/or antifungal susceptibility results.