Prior use of fluconazole is a modifiable risk factor for the isolation of fluconazole-nonsusceptible Candida species. Optimization of the use of fluconazole by appropriate dose or duration may be able to minimize the risk of resistance. The objective of this study was to evaluate the effects of prior fluconazole therapy, including the dose and duration, on fluconazole susceptibility among Candida species isolated from hospitalized patients with candidemia. A retrospective cohort study of hospitalized patients with a first occurrence of nosocomial candidemia, from 2006 to 2009, was carried out. The relationships between the initial dose and duration of prior fluconazole therapy and the isolation of fluconazole-nonsusceptible Candida species were assessed. An initial fluconazole dose greater than 2 mg/kg and less than 6 mg/kg of body weight was considered suboptimal. A total of 177 patients were identified, of whom 133 patients aged 61 ± 16 years (56% male, 51% Caucasian, 51% with an APACHE II score of ≥ 15) had candidemia more than 2 days after the hospital admission day. Nine of 107 (8%) patients with fluconazole-susceptible Candida species and 9 of 26 (35%) patients with fluconazole-nonsusceptible Candida species had prior fluconazole exposure (risk ratio [RR], 3.03; 95% confidence interval [95% CI], 1.57 to 5.86; P, 0.0022). Preexposure with an initial dose of fluconazole greater than 2 mg/kg and less than 6 mg/kg occurred in 3 of 9 (33%) and 8 of 9 (89%) patients with fluconazole-susceptible and fluconazole-nonsusceptible Candida species, respectively (P, 0.0498). We conclude that patients with candidemia due to fluconazole-nonsusceptible Candida species were more likely to have received prior fluconazole therapy. Suboptimal initial dosing of prior fluconazole therapy was associated with candidemia with fluconazole-nonsusceptible Candida species.