Abstract Objective. Complex renal cysts represent a major clinical problem, since it is often difficult to exclude malignancy. The Bosniak classification system, based on computed tomography (CT), is widely used to categorize cystic renal lesions. The aim of this study was to evaluate critically available data on the Bosniak classification. Material and methods. All publications from an Entrez Pubmed search were reviewed, focusing on clinical applicability and the use of imaging modalities other than CT to categorize complex renal cysts. Results. Fifteen retrospective studies were found. Most series were small, with only six studies comprising more than 30 patients. When the results of these six studies were pooled the percentage of malignancy was: category I, 0%; category II, 15.6%; category IIF, 0%; category III, 65.3%; and category IV, 91.7%. Category IIF was included in only three studies. Both magnetic resonance imaging and contrast-enhanced ultrasonography (US) seemed to have reasonable agreement with CT, although both modalities had a tendency to upgrade lesions. Conclusions. The main problem with the Bosniak classification is in separating category II and III lesions, which is of major importance, since decisions on intervention are based on this separation. The use of category IIF seems promising to detect those category II lesions that eventually will develop into malignancy and reduce overtreatment of lesions originally classified as category III. The scientific basis of this "new" classification strategy is, however, still missing. Data on other imaging modalities are too limited for conclusions to be drawn.