The effect of intravenous versus intranasal TRH stimulation was compared in geriatric patients. In patients receiving both iv and nasal tests (N = 35) there was a good correlation between the TSH responses, but in 3 cases the suspicion of inadequate nasal TRH effect arose. The coefficient of correlation of basal to delta TSH was better in iv tested patients (116 patients in each group, one half having a positive the other half a negative TRH test). In the majority of patients with suspicious incongruity of basal and delta TSH the nasal test was done. The specificity and sensitivity of various basal TSH 'cut-off' points to predict a positive TRH test were better in patients with iv TRH tests (in each group 96 consecutively admitted patients). Insufficiency of the nasal test in geriatric patients is mainly explained by the inability of the old people to aspirate the nasal spray effectively. Besides the advantages of iv TRH application in geriatric patients, the frequency of adverse reactions (14% versus 0%) must be considered. Thus, in a hospitalized geriatric patient, the TRH test should be performed iv in the recumbent position; however, for examination of geriatric outpatients the nasal test is recommended.