Hemoglobin A from adult humans does not contain any coded isoleucine}, thus its presence in hemoglobulin A must originate through errors in transcription or translation, or by somatic mutations arising during DNA replication. Errors in transcription occur infrequently to form altered mRNA, tRNA and rRNA; such errors change the coding in the mRNA and may reduce the fidelity of the tRNA as regards both the kind of amino acid it accepts and the mRNA codons it recognizes-. Errors in translation arise through the attachment of wrong amino acids to tRNA (amino acyl synthetase errors) and the imprecise recognition of the mRNA codons by the tRNA anticodons (translational variation). Somatic mutations result from mistakes in replication of DNA; many single base substitution mutations change nonisoleucine into isoleucine codons? and these "mutant" cells could have hemoglobin mRNA with isoleucine codons. For this reason, an increase in the quantity of isoleucine in hemoglobin A would be expected in humans who have been exposed to agents that cause base substitution mutations. Radiation may cause base substitution mutations in human somatic cells but this possibility has never been established. This report describes the analysis of the isoleucine content of hemoglobin A from 13 Marshallese who were exposed to fallout from an atomic bomb test in March of 1954 compared with hemoglobin A from 12 Marshallese controls. Methods The blood was collected by Dr. Conard in March of 1974 and received in Oak Ridge on April 4, 1974. Hemoglobin was purified initially by molecular seiving on Sephadex G-220 and then by three successive chromatographic separations of the carbonmonoxyhemoglobin, methemoglobin, and metcyanhemoglobin