The percent of the eligible population examined on at least one occasion during the three year period was: Rongelap Utirik Comparison 97% 100% 94%, These figures do not include several persons residing outside the Marshall Islands. Most exposed personsin this category have medical examinations arranged through a local physician by the Department of Energy or the Marshall Islands Medical Program. The acceptance rate for mammography among eligible women was 100%. For sigmoidoscopy, about 50%of ageeligible persons elect to undergo this procedure on a regular basis. MEDICAL FINDINGS Overall Survival: After thirty-three years there continues to be no significant difference in the survival curves of the high-exposure Rongelap group, the lowexposure Utirik group, and the unexposed Rongelap population followed for the purpose of comparison (Fig. 1). Estimates of the survival distribution by the actuarial life table method were analyzed by Mantel-Cox and Breslow statistics for testing the equality of the survival curves. The “p” values were 0.68 by both techniques. In the Brookhaven National Laboratory report covering January 1983 through December 1984, it was noted that Okajima etal. (1985) suggested that medical programs providing health screening might lead to an underestimation of the effect of radiation on mortality. In particular, it was postulated that this could explain the lower age-specific death rates from all causes among Nagasaki A-bomb survivors, comparedto a control population. The effect of medical examinations on the survival of the exposed Marshallese is unknown. On the one hand about 15 percent of the Comparison group selected in 1957 is no longer seen because those individuals have voluntarily foregone examination. In addition, BNLreferrals for the Comparison group are channeled into the Marshallese Health Services system, whereas selected medical problems in the exposed groups can be referred directly to tertiary care facilities in the United sates other hand, the exposed populations of RonBelap and Utirik have received equivalent medical attention from the BNL program since 1972, and yet, despite the far higher radiation dose received by the Rongelap group, the survival curves are similar. Anotherfactor that contributes to the diffi- culty in interpreting differences in the group survivals in Fig. 1 is that the population used to construct the “Rongelap unexposed” curve was selected in 1957, and it is in that year that their survival is graphed as one-hundred percent; ie., data from three years of observation, during which some deaths occurred, had already been acquired from the two exposed populations. Causes of Recent Mortality: The number of deaths occurring in thelast three yearsare as follows: Rongelap exposed - 2; Utirik exposed - 9; Comparison group - 10, The specific clinical situations are described below. Rongelap Subject No. 1. The causes of death listed on the death certificate of this 81-year-old woman in June 1985 were “Inanition” and “Senility.” When seen in March 1985, she had a normal blood pressure and cardiac examination revealed “premature beats." In 1984 she was noted to have cataracts,atrial fibrillation, and complaints of urinary incontinence, some cough, constipation, and joint pains. Her hemoglobin was 12.7 g/dl, the mean corpuscular volume was 92 fl, and the white blood cell count was 6,600per ul with a normal differential. Subject No. 11. This 81-year-old man died in 1987 of unknown cause. Diagnoses madeduring the preceding four years included severe osteoarthritis, chronic obstructive pulmonary disease with bullous emphysema, macrocytic anemia that was being treated with vitamin B12 injections, cataracts, and “organic brain syndrome.” He had declined a medical examination when visited at his home in September 1986, but did not appear acutelyill at that time. Utirik Subject No. 2123. This 47-year-old man died in December 1986 from biopsy-proven hepatocellular carcinoma. His alpha fetoprotein level was elevated and the serum contained hepatitis Bsurface antigen but no delta antibody. No evidence of tumor was found at his March 1986 examination. Symptoms related to the tumor developed in June of that year.