Data on human monkeypox collected in Zaire during the six years 1981-86 were analysed to assess the extent of interhuman transmission of monkeypox virus. Among the 2278 persons who had close contact with 245 monkeypox patients infected from an animal source, 93 fell ill and were presumed to have been infected from the known human source: 69 of these were spread in the first generation, 19 in the second generation, and the remaining five cases in the third and fourth generation.The secondary attack rates were correlated with the age, sex, place of residence, and vaccination status of the contacts. There was an overall 3% probability of becoming ill following infection from a known human source. The affected household was the main focal point for interhuman transmission of monkeypox virus. The highest attack rate (11.7%) occurred among unvaccinated household contacts in the age group 0-4 years. However, the majority of susceptible persons who had been close to patients in the confined space of poorly ventilated huts failed to develop illness. There was no evidence of an increase in the secondary attack rate between 1970-80 and 1981-86.The inefficient spread from person to person, even in conditions of maximum exposure, supports the concept that monkeypox virus is poorly adapted for sustained transmission between humans and that such transmission does not pose a significant health problem.