Introduction: The frequency of invasive mycoses has increased dramatically during the past two decades owing to medical advances such as intensive cancer chemotherapy, broad-spectrum antimicrobial therapy, invasive medical devices, organ transplantation, human immunodeficiency virus (HIV) disease epidemic and an expanding aging population. There were few Indian studies regarding the incidence and risk factors for candidemia. Hence the aim of this work was to evaluate the changes in the prevalence of candidemia and invasive candidiasis in a tertiary care hospital and also to assess the risk factors and predictors of mortality Materials and Methods: Nonsystematic review of patients with candidemia/invasive candidiasis was done during the period 1999 to 2004. All in-patients who had shown signs and symptoms of nosocomial blood stream infection were screened for candidial infection. Among these, 29 patients had candidemia/invasive candidiasis. Demographic and clinical data of these patients were recorded on a standardized form, which included age, sex, site of isolation, infectious diagnosis, underlying conditions, predisposing factors, catheter status and clinical outcome. The data were collected during the years of 1999 to 2004, which is divided into two time periods (1999-2001 and 2002-2004). Data collected during these different time spans are compared with each other. Results: A total of 255 patients were screened during the study period. Among these, 100 patients were screened during the period 1999-2001 and 155 patients were screened during the year 2002-2004. Out of these patients, 29 showed positive cultures in blood or other sterile site (ascitic fluid, bronchial aspirate and urine from suprapubic puncture). Out of these, 24 were males and five were females. The most common risk factor was use of intravenous canulae (62.1%), followed by prolonged use of antibiotics (34.5%) and HIV infection (24.1%). There were no statistically significant differences in the risk factors during the two different study periods. Candida was mainly isolated from blood (75.9%). Other sources included ascitic fluid (10.4%), bronchial aspirate (3.4%), sputum (3.4%) and urine (6.9%). Distributions of sources were comparable during the two study periods. Candida albicans, Candida tropicalis and Candida parapsilosis caused 89.7%, 3.4%, 6.9% of the candidemia episodes respectively. The overall mortality was 51.7%. Conclusion: The present study emphasizes the importance of candidemia among hospitalized patients. Continued surveillance of candidemia will be important to track trends of this serious infection and to document changes in its epidemiological features. More active screening in high-risk groups should be done to avoid diagnostic delay. Risk factors like prolonged use of multiple antibiotics, central venous catheters, mechanical ventilation and prolonged hospital stay should be restricted whenever possible. Timely use of antiretroviral drugs and other measures to improve the immunity of HIV patients may help to decrease the incidence of candidemia in this patient population.