Pharmacological treatment of bipolar affective disorder was examined by case note review of 120 subjects under the care of four hospitals in the North-East of England and compared with that recommended in published consensus opinions and also with the findings of previous surveys. Although the only comparison available is with other geographical regions, it appears that lithium use has remained relatively constant in recent years (53% in this sample). The use of anticonvulsants (62%) was greater than in previous studies, as was use of mood stabilizers in combination (37% overall, 53% in rapid cycling subjects). Antidepressant use was found in 23% of subjects. Eleven percent of subjects taking antidepressants were not prescribed a mood stabilizer and 43% of antidepressants prescribed were tricyclics, possibly increasing the risk of switching to mania. Fifteen percent had no mood stabilizer and antipsychotic use was almost equally split between typical and atypical drugs. Practice in the population studied is nearer than treatment described in previous work to that suggested in the literature. This may result from either, or most probably a combination of, the local availability of a specialist mood disorder service and evolution of overall prescribing practice. However, it still falls short in a number of respects of such recommendations as exist and this requires further examination and monitoring.