The performance of 222 lithium assays was studied in a psychiatric hospital to determine if their use was appropriate and therapeutically beneficial. Charts of all patients whose serum lithium levels were measured in a month were reviewed to determine whether the assays were indicated, whether they were performed correctly, and whether the dosage was adjusted appropriately. Of the assays reviewed, 16 percent were performed for inappropriate indications and 98 percent were drawn at correct times. Of 157 assays that should not have resulted in a dosage change, 8 changes were made; of 63 assays that should have resulted in a dosage change, only 23 changes were made. Inappropriate orders for assays cost the hospital +5016 per year. If routine serum lithium assay monitoring is changed to a three-month frequency, the total cost savings would be +16,632 per year. Clinical pharmacists with specialized knowledge of clinical pharmacokinetics should work with physicians and treatment teams in monitoring serum drug assays to ensure appropriateness of assay request, more complete interpretation of results, and proper follow up.