Eleven patients with homocystinuria were given a trial of treatment with high doses of oralpyridoxine. Biochemically 6 patients responded, as shown by the return of their plasma amino acid pattern to normal or near normal. 5 showed no response. The same type of biochemical response was recorded in all affected members of a family. It is suggested that in homocystinuria two different genetically determined basic defects exist in the cystathionine synthetase system, producing the same biochemical and clinical picture. In those patients who are biochemically unresponsive to pyridoxine, the defect appears to be a deficiency of the apoenzyme, cystathionine synthetase. On the other hand, in patients who respond biochemically to pyridoxine, the disorder seems to represent a further example of a pyridoxine dependency syndrome. It is too early to say if treatment with pyridoxine will prevent the sequelae of this disorder, and only by prolonged follow-up of the younger patients will this become known. The possibility of a folic acid deficiency arising while on treatment is stressed. The work was supported by a grant from the Medical Research Council, London.