We conducted laboratory tests to assess the sensitivity to the insect repellent 1-piperidinecarboxylic acid, 2-(2-hydroxyethyl)-, 1-methylpropylester (known as KBR 3023 or Picaridin, trade name Bayrepel) of West African strains of the yellow fever mosquito Aedes aegypti and of malaria vectors of the Anopheles gambiae complex, in comparison with the standard repellent N,N-diethyl-3-methyl-benzamide (DEET). Test mosquitoes were exposed according to a 'separate arms' protocol to logarithmic dose increments applied on one arm of human subjects to evaluate the relative potency, and the median effective dosages (ED50 and ED90). According to a logistic regression model fitted to the experimental data, the dose-response relationship for the two repellents was the same within each species, thus pooled ED values were assessed for each mosquito separately. The median ED of KBR 3023 and DEET was estimated at 0.78 (95% confidence limits (CI): 0.57-1.04) and at 0.018 microg/cm2 (0.004-0.052) for mosquitoes of the An. gambiae complex and Ae. aegypti, respectively. ED90 values were 125.6 (81.4-201.3) and 24.0 microg/cm2 (5.7-208.5) for An. gambiae s.l. and Ae. aegypti, respectively. The relative potency of KBR 3023 was not significantly different from that of DEET for An. gambiae s.l. (95% confidence limits 0.7-1.0), whereas in the case of Ae. aegypti it was with 95% probability 1.1-2.0 times more potent than DEET. On the basis of available evidence, KBR 3023 represents a promising alternative to DEET for personal protection against bites of these important vectors of disease in the Afrotropical region.