Laboratory-bred Culex quinquefasciatus were fed on carriers with low and moderate densities of microfilariae (mf) of Wuchereria bancrofti. In the first series of experiments, mosquitoes were dissected 12 d after feeding. The percentage of infected mosquitoes and the numbers of larvae per infected mosquito were directly proportional to the mf density at the time of feeding. There was an overall high infection rate and a wide distribution of larvae per individual mosquito at all levels, except for the 4 lowest counts. Of the 4 carriers with counts of 5 mf/ml and less, 3 were capable of infecting Cx quinquefasciatus, giving infection rates of 1.0%, 7.4% and 12.0% respectively. In the second series, some mosquitoes were dissected immediately after feeding and the remainder 12 d later. There was a good correlation between the number of mf ingested and the number of infective larvae per mosquito. The high infection rates in Cx quinquefasciatus when fed on low-density microfilaraemia carriers, and the varying number of larvae in individual mosquitoes, suggest that low-density carriers could be a source of infection. Field studies were also carried out in 3 different area with mf rates of 7.24%, 0.72% and 0.16%, respectively. In the 2 areas with low mf rates, infection rates in mosquitoes were 1.32% and 1.08% respectively. Cx quinquefasciatus fed on a carrier with a residual microfilaraemia of 19 mf/ml following treatment with diethylcarbamazine had an infection rate of 13.8%. These studies suggest that the examination of recently fed house-resting populations of Cx quinquefasciatus could be a sensitive method for measuring the prevalence of mf in the human population.