Summary A good but not absolute correlation between the presence of the gastric parietal cell antibody and impairment of gastric secretion was observed in the relatives of patients with pernicious anaemia. By analogy, although the majority of relatives without the antibody secreted normal amounts of acid and of intrinsic factor nearly one‐fifth secreted unexpectedly small amounts of intrinsic factor.There was a highly significant correlation between the amounts of acid and of intrinsic factor secreted in response to histamine or gastrin and as a rule impairment of secretion affected these substances to the same extent. However, one young relative who did not have gastric or thyroid antibodies was considered to have selective impairment of intrinsic factor secretion.Thirty‐eight subjects had direct measurements of intrinsic factor and tests of vitamin B12 absorption. Twenty‐seven of these secreted less than 600 units of intrinsic factor in the hour following stimulation and all but two had malabsorption of radioactive vitamin B12. Similarly all but two of the 11 subjects who secreted more than this amount had normal absorption of the vitamin.Twenty‐two per cent of the probands of the study had a family history of pernicious anaemia, 18 per cent one of thyroid disease, 20 per cent one of diabetes mellitus and 3.8 per cent had a family history of carcinoma of the stomach.Analysis of the distribution of inherited characteristics showed that there was a slight but not significant excess of blood group A in the patients with pernicious anaemia and that the proportion of patients with Lea‐positive red cells was lower than the figure quoted for the general population. Neither in the probands nor in their relatives was there an association between blood group A or the Lea group and the tendency to form circulating antibodies to gastric parietal cells. Moreover, although the proportion of patients with pernicious anaemia who had light‐coloured eyes was of the same order as in a previous study from this department, the prevalence of gastric parietal cell antibodies was no higher in such patients than in those with dark eyes, nor was the prevalence of the antibodies significantly increased in patients who gave a history of premature greying.The gastric parietal cell antibody was no more common in the spouses of probands than in matched control subjects. It was further equally common in relatives of probands without the antibody as in relatives of probands with the antibody. Analysis of three families in which both parents had pernicious anaemia and of seven families in which it was possible to study both parents of the proband suggested that the inheritance of the tendency to form gastric parietal cell antibodies was not a simple one. It appeared that the varying patterns of distribution of the antibody in these families could be best explained by suggesting that the tendency to form the antibody is inherited as a dominant characteristic with an incomplete degree of penetrance.