Human infectiousness to mosquitoes can be estimated by 2 tests: direct feeding on the skin and membrane feeding on venous blood. To validate the membrane feeding assay, the infectiousness of Plasmodium falciparum gametocyte carriers to Anopheles gambiae was estimated by these 2 methods in the same individuals in a rural area of Cameroon. Results from 37 experiments showed that direct feeding gave significantly higher infection rates than membrane feeding. We observed an average of 19·4% infected mosquitoes by direct feeding compared with 12·1% by membrane feeding, and a mean oocyst load of 5·63 by direct feeding compared with 2·65 by membrane feeding. However, there was a very good concordance between the 2 tests: 84·3% with the Kappa test on percentages of infected mosquitoes and 98·7% with the interclass correlation coefficient on oocyst loads. In addition, we found a good linear correlation between the 2 methods.