Conflicting opinions on the value of a skin test in the diagnosis of human filariasis emphasized the need for a careful examination of this procedure. The evaluation was made by asking workers in different countries to use an antigen prepared from Dirofilaria immitis in groups of people who could be examined for parasitic infection.In one non-endemic area, repeated tests over a 1-year period did not lead to sensitization, but the reactions of individuals varied from test to test. In endemic areas of filariasis, exposure to infective bites seemed to influence the pattern of skin reactions to a greater degree than did the development of overt infection with Wuchereria bancrofti. No particular size of reaction could be considered indicative of filarial infection.THE VALUE OF THE SKIN TEST WITH THIS ANTIGEN SEEMS LIMITED TO TWO SITUATIONS: (1) a large reaction may help to confirm a clinical diagnosis of filariasis when parasites cannot be found, and (2) the frequency distribution of skin reaction sizes in local populations may help, where blood surveys are impossible, to indicate areas in which some filarial infection is being transmitted.Results of skin test surveys should be expressed as frequency distributions of reaction sizes and negative/positive classification should be avoided.