BACKGROUND The potential adverse central nervous system effects of H1-receptor antagonists have not been optimally studied in the elderly. OBJECTIVE We hypothesized that newer H1-receptor antagonists such as cetirizine and loratadine would cause less central nervous system dysfunction than the older H1-receptor antagonists diphenhydramine and chlorpheniramine in this population, as they do in younger subjects. METHODS We performed a randomized, double-blind, single-dose, placebo-controlled, 5-way crossover study in 15 healthy elderly subjects (mean age 71 +/- SD 5 years). On study days at least 1 week apart, they received cetirizine 10 mg, loratadine 10 mg, diphenhydramine 50 mg, chlorpheniramine 8 mg, or placebo. Outcome measures, recorded before and 2 to 2.5 hours after dosing were latency of the P300 event-related potential in which increased latency reflects a decreased rate of cognitive processing, visual analogue scale for subjective somnolence, and histamine skin tests for measurement of peripheral H1-blockade. RESULTS The changes in P300 following each treatment yielded variances that were not equal (P > .05), precluding usual statistical analysis of the means. These variances were ranked: chlorpheniramine > diphenhydramine > loratadine > placebo > cetirizine. The rank of mean differences in the visual analogue scale increase from pre-dose baseline was: diphenhydramine > chlorpheniramine > cetirizine > loratadine > placebo. All H1-receptor antagonists suppressed the histamine-induced wheal and flare significantly compared to baseline. CONCLUSIONS In the elderly, the new H1-receptor antagonists cetirizine and loratadine are less likely to cause adverse central nervous system effects than the old H1-antagonists chlorpheniramine or diphenhydramine, but this requires confirmation using additional objective tests of central nervous system function.