Hungry Anopheles funestus females were caught in CDC light traps positioned overnight in village houses in which the occupants were protected by bed nets. The sampling was performed in six huts near Muheza, in North-eastern Tanzania, September to November 1972. Parous, unfed mosquitoes were carefully examined, usually under phase-contrast optics, for the presence of developing Wuchereria bancrofti larvae. Because A. funestus seek blood meals at three-day intervals, most filariae were three, six, nine, 12 and 15 days old, and could with confidence be assigned to one of these five easily recognized age groups. Of 3,289 unfed A. funestus examined, 1,723 (52%) were parous. Developing and mature infections amongst the parous were 272 in number. Simultaneous infections of two age groups numbered 21, one mosquito was triply infected, thus there were 249 infected mosquitoes (7·57% of the total). The regression of loge number of infections ( y) on estimated age in days ( X), agreed with an hypothesis of linearity (r = −0·9967); the equation log e y = 5·25− 0·1646X suggested an average daily mortality rate of 16·46% in good agreement with the mean 15-6% observed in A. funestus by Gillies & Wilkes. The equation above affords an estimate of the probability of an A. funestus becoming infected at a single random feeding. The expected number of day 0 infections, given by the antilog of the first term of the equation and taken as a fraction of the biting population, gives the estimate. The values obtained were 0-058-0-087. Use of Macdonald's 1952 formula for the sporozoite rate provided estimates of 0-071-0-081. Age-specific worm densities in A. funestus did not vary significantly between day three and day 15 infections, and no evidence was obtained to suggest that the higher worm burdens influenced the survival of their hosts. The frequency of superinfections did not support the contention that a previously established infection inhibits the chance of a second, younger infection from becoming established.