This paper reports a 15-year follow-up study of 12,693 persons in Linxian, China, who were originally screened by esophageal balloon cytology in 1974. The purpose of the study was to evaluate the ability of esophageal balloon cytology to identify individuals at increased risk for developing esophageal cancer. Age, sex and cytologic diagnoses were obtained from the original 1974 records, and information on vital status, cancer experience and potential confounding risk factors was collected from interviews and medical abstracts in 1989. A total of 1,162 incident cases of esophageal cancer and 993 deaths due to esophageal cancer were identified and used in this analysis. The follow-up study showed that the risk of esophageal cancer incidence and mortality increased in parallel with the presumed severity of the 1974 Chinese cytologic diagnoses. After adjusting for potential confounding factors, the relative risks (and 95% confidence intervals) for esophageal cancer incidence, by cytologic diagnosis, were: normal, 1.00; esophagitis, 1.52 (1.07-2.14); hyperplasia, 1.17 (1.02-1.33); dysplasia 1, 1.53 (1.10-2.14); dysplasia 2, 1.89 (1.47-2.41); and suspicious for cancer, 5.77 (3.79-8.80). These results suggest that esophageal balloon cytology, as performed and interpreted in Linxian in 1974, successfully identified individuals at increased risk for esophageal cancer.