Gout has long been thought to be hereditary and several genetic hypotheses have been proposed for the occurrence of hyperuricaemia (Smyth, Cotterman, and Freyberg, 1948; Stecher, Hersh, and Solomon, 1949; Hauge and Harvald, 1955). Blumberg (1965), in a critical review of these studies, has concluded that they are far from ideal in testing a genetic hypothesis, and comments that this may be why such studies have provided inconsistent results. Indeed, recently, O'Brien, Burch, and Bunim (1966), in a carefully controlled epidemiological study of the Blackfeet Indians in Montana and the Pima Indians in Arizona, showed that the trait of hyperuricaemia "is for the most part not a genetic one but is instead largely determined by environmental factors". These authors, did, however, note some familial aggregation of hyperuricaemia, but genetic analysis was not consistent with either a simple recessive or dominant inheritance; analysis of a polygenic trait indicated a rather low heritability. It occurred to us that it might be of interest, in view of the present controversy surrounding the genetics of gout and hyperuricaemia, to ascertain the strength of the genetic and environmental components in the control of the serum uric level in normouricaemic subjects. To do this we have adopted the twin study method where the mean intrapair variance of serum uric acid levels observed in monozygotic (MZ) (genetically-identical) twin pairs has been contrasted with the mean intrapair variance of serum uric acid levels in dizygotic (DZ) (genetically non-identical) twin pairs and the results of this comparison have been analysed by calculating the variance ratio and assessing its significance on the F ratio, as has been described by Osborne and De George (1959).