An investigation of some of the clinical associations of low serum folate and B12 values found in a previous survey of 432 psychiatric inpatients is presented. There were 105 (22.5 per cent) “low folate” (12” (12, and 44.1 per cent of low B12 patients had received drugs during the three preadmission weeks; barbiturates, phenothiazines, and antidepressants (in that order) were the commonest drugs among low folate patients, and phenothiazines among low B12 patients. Significantly more low B12 patients (38.2 per cent) than low folate patients (18.6 per cent) were physically ill. Many low folate (61.6 per cent) and low B12 (47.1 per cent) patients had been ill for more than 1 year, and serious malnutrition was frequent among both low folate (22.1 per cent) and low B12 (32.4 per cent) groups. Low serum levels were infrequently associated with pregnancy and gastrectomy, and no case of intestinal malabsorption was seen. These extrinsic factors were not found in only 21 low folate and 14 low B12 patients (“primary” or “idiopathic” group) among whom no particular mode of psychiatric presentation was seen. Low folate patients had significantly more hematological abnormalities than normal folate patients, whereas low B12 compared with normal B12 patients did not. A retrospective survey of the results of treatment with folic acid and vitamin B12 supplements was carried out because it was found that they had been given to patients with low values in such a haphazard way as to be distributed virtually randomly. Among the low folate patients, 39 received folate and 63 did not. Folate-treated patients with organic psychoses, endogenous depression, and schizophrenia, when assessed on a 4-point scale, were found to be in a better clinical state at discharge than untreated patients; and their stay in the hospital was significantly shorter. There was no evidence that cyanocobalamin-treated low B12 patients were in a better clinical state at discharge or spent a shorter time in the hospital than similar untreated patients, though the assessment period was probably too short for firm conclusions to be drawn.