Mortality due to hematological tumors in towns near Spain's seven nuclear power plants and five nuclear fuel facilities during the period 1975-1993 was ascertained. The study was based on 610 leukemia-, 198 lymphoma-, and 122 myeloma-induced deaths in 489 towns situated within a 30-km radius of such installations. As control areas, we used 477 towns lying within a 50- to 100-km radius of each installation, matched by population size and a series of sociodemographic characteristics (income level, proportion of active population engaged in farming, proportion of unemployed, percentage of illiteracy, and province). Relative risk (RR) for each area and the trends in risk with increasing proximity to an installation were analyzed using log-linear models. None of the nuclear power plants registered an excess risk of leukemia-induced mortality in any of the surrounding areas. Excess risk of leukemia mortality was, however, observed in the vicinity of the uranium-processing facilities in Andújar [RR, 1.30; 95% confidence interval, 1.03-1.64] and Ciudad Rodrigo (RR, 1.68; 95% confidence interval, 0.92-3.08). Excess risk of multiplemyeloma mortality was found in the area surrounding the Zorita nuclear power plant. Statistical testing revealed that, with the single exception of multiple myeloma, none of the tumors studied showed evidence of a rise in risk with proximity to an installation. No study area yielded evidence of a raised risk of leukemia mortality among persons under the age of 25 years. More specific studies are called for in areas near installations that have been fully operational for longer periods. In this connection, stress should be laid on the importance of using dosimetric information in all future studies.