The authors used competitive protein binding assay and radioimmunoassay to measure cortisol levels in 38 normal control subjects three times before and three times after administration of 1 mg of dexamethasone. They found significant interassay differences at 11:00 p.m. before dexamethasone and at all three postdexamethasone times. Analysis of variance revealed significant overall positive relationships between age and cortisol levels measured by both techniques. Age correlated significantly with postdexamethasone cortisol levels measured by radioimmunoassay but not when measured by competitive protein binding assay. Clinicians should obtain data from their laboratories as to appropriate cutoffs for cortisol suppression on the specific assay used.