Routes of asbestos exposure consist of occupational and non-occupational exposures, and furthermore the latter is classified as para-occupational, neighborhood or true general environmental exposure. Consequently, in order to evaluate health risk caused by neighborhood exposure to asbestos, it is necessary to exclude risk due to the other exposure routes from overall risk. We reviewed epidemiological studies on the relationship between neighborhood asbestos exposure and risk of mesothelioma. In studies on a crocidolite mine in South Africa and a chrysotile mine in Canada, occupational exposure was not excluded. In studies on a crocidolite mine in Australia and an asbestos manufacturing factory in U.S.A., risk caused by non-occupational exposure was evaluated, but the risk was not classified as para-occupational and neighborhood exposures. In a study on an asbestos cement factory in Italy, first, occupational and para-occupational exposures were excluded, and next, the incidence rate of mesothelioma in neighborhood residents was calculated, so that risk caused by neighborhood exposure could be evaluated. In case-control studies in Italy, South Africa, three European countries and the U.K., risks caused by occupational, para-occupational and neighborhood exposures were evaluated separately. As a whole, relative risk (RR) of neighborhood exposure in crocidolite and amosite mines was about 10 to 30 and RR in major asbestos factories was about 5 to 20. On the other hand, statistically significant RR of neighborhood exposure was not observed in chrysotile mines and some asbestos facilities.