Two cohorts of Chernobyl cleanup workers from Estonia (4,786 men) and Latvia (5,546 men) were followed from 1986 to 1998 to investigate cancer incidence among persons exposed to ionizing radiation from the Chernobyl accident. Each cohort was identified from various independent sources and followed using nationwide population and mortality registries. Cancers were ascertained by linkage with nationwide cancer registries. Overall, 75 incident cancers were identified in the Estonian cohort and 80 in the Latvian cohort. The combined-cohort standardized incidence ratio (SIR) for all cancers was 1.15 (95% confidence interval (CI) = 0.98-1.34) and for leukemia, 1.53 (95% CI = 0.62-3.17; n = 7). Statistically significant excess cases of thyroid (SIR = 7.06, 95% CI = 2.84-14.55; n = 7) and brain cancer (SIR = 2.14, 95% CI = 1.07-3.83; n = 11) were found, mainly based on Latvian data. However, there was no evidence of a dose response for any of these sites, and the relationship to radiation exposure remains to be established. Excess of thyroid cancer cases observed may have been due to screening, the leukemia cases included 2 unconfirmed diagnoses, and the excess cases of brain tumors may have been a chance finding. There was an indication of increased risk associated with early entry to the Chernobyl area and late follow-up, though not statistically significant. Further follow-up of Chernobyl cleanup workers is warranted to clarify the possible health effects of radiation exposure.