Non-albicans Candida (NAC) species cause 35-65% of all candidaemias in the general patient population. They occur more frequently in cancer patients, mainly in those with haematological malignancies and bone marrow transplant (BMT) recipients (40-70%), but are less common among intensive care unit (ITU) and surgical patients (35-55%), children (1-35%) or HIV-positive patients (0-33%). The proportion of NAC species among Candida species is increasing: over the two decades to 1990, NAC represented 10-40% of all candidaemias. In contrast, in 1991-1998, they represented 35-65% of all candidaemias. The most common NAC species are C. parapsilosis (20-40% of all Candida species), C. tropicalis (10-30%), C. krusei (10-35%) and C. glabrata (5-40%). Although these four are the most common, at least two other species are emerging: C. lusitaniae causing 2-8% of infections, and C. guilliermondii causing 1-5%. Other NAC species, such as C. rugosa, C. kefyr, C. stellatoidea, C. norvegensis and C. famata are rare, accounting for less than 1% of fungaemias in man. In terms of virulence and pathogenicity, some NAC species appear to be of lower virulence in animal models, yet behave with equal or greater virulence in man, when comparison is made with C. albicans. Mortality due to NAC species is similar to C. albicans, ranging from 15% to 35%. However, there are differences in both overall and attributable mortality among species: the lowest mortality is associated with C. parapsilosis, the highest with C. tropicalis and C. glabrata (40-70%). Other NAC species including C. krusei are associated with similar overall mortality to C. albicans (20-40%). Mortality in NAC species appears to be highest in ITU and surgical patients, and somewhat lower in cancer patients, children and HIV-positive patients. There is no difference between overall and attributable mortality, with the exception of C. glabrata which tends to infect immunocompromised individuals. While the crude mortality is low, attributable mortality (fungaemia-associated mortality) is higher than with C. albicans. There are several specific risk factors for particular NAC species: C. parapsilosis is related to foreign body insertion, neonates and hyperalimentation; C. krusei to azole prophylaxis and along with C. tropicalis to neutropenia and BMT; C. glabrata to azole prophylaxis, surgery and urinary or vascular catheters; C. lusitaniae and C. guilliermondii to previous polyene (amphotericin B or nystatin) use; and C. rugosa to burns. Antifungal susceptibility varies significantly in contrast to C. albicans: some NAC species are inherently or secondarily resistant to fluconazole; for example, 75% of C. krusei isolates, 35% of C. glabrata, 10-25% of C. tropicalis and C. lusitaniae. Amphotericin B resistance is also seen in a small proportion: 5-20% of C. lusitaniae and C. rugosa, 10-15% of C. krusei and 5-10% of C. guilliermondii. Other NAC species are akin to C. albicans-susceptible to both azoles and polyenes (C. parapsilosis, the majority of C. guilliermondii strains and C. tropicalis). Therefore, 'species directed' therapy should be administered for fungaemia according to the species identified-amphotericin B for C. krusei and C. glabrata, fluconazole for other species, including polyene-resistant or tolerant Candida species (C. lusitaniae, C. guilliermondii). In vitro susceptibility testing should be performed for most species of NAC in addition to removal of any foreign body to optimize management.