The aim of the present study was to evaluate serum concentrations of vascular endothelial growth factor (VEGF) in patients with vulvar cancer and healthy female controls with respect to correlation of VEGF with clinicopathological parameters and impact on the patients' prognosis. Serum concentrations of VEGF were measured using a commercially available ELISA. Results were correlated to clinical data. Median serum concentrations of VEGF in patients with vulvar cancer (n = 41) and healthy female controls (n = 130) were 260 (range, 33-1216) pg/ml and 216 (range, 0-777) pg/ml, respectively (Mann-Whitney U test, P = 0.048). Serum concentrations of VEGF significantly correlated with tumor stage (Mann-Whitney U test, P = 0.02) but not with histological grade (Mann-Whitney U test, P = 0.2). In a univariate analysis, elevated pretreatment serum concentrations of VEGF were significantly correlated with a shortened disease-free and overall survival (Wilcoxon test, P = 0.03; and Wilcoxon test, P = 0.04, respectively). A multivariate Cox regression model considering tumor stage and serum concentrations of VEGF revealed, however, that serum concentrations of VEGF did not confer additional prognostic information to that already obtained by the established prognosticator tumor stage (multivariate Cox regression model: P = 0.9 and P = 0.8, respectively). Our data indicate that angiogenesis, as reflected by serum concentrations of VEGF, plays a functional role in vulvar carcinogenesis. VEGF seems to be a mediator of vulvar tumor growth but not of tumor cell dedifferentiation. Although associated with impaired disease-free and overall survival, pretreatment serum concentrations of VEGF are not an independent predictor of outcome in patients with vulvar cancer.