Recent in vitro studies have suggested that P‐glycoprotein (Pgp) and passive membrane permeability may influence the brain concentrations of non‐sedating (second‐generation) antihistamines. The purpose of this study was to determine the importance of Pgp‐mediated efflux on the in vivo brain distribution of the non‐sedating antihistamine cetirizine (Zyrtec®), and the structurally related sedating (first‐generation) antihistamine hydroxyzine (Atarax®). In vitro MDR1‐MDCKII monolayer efflux assays demonstrated that cetirizine was a Pgp substrate (B → A/A → B + GF120918 ratio = 5.47) with low/moderate passive permeability (Papp B → A = 56.5nm/s). In vivo, the cetirizine brain‐to‐free plasma concentration ratios (0.367 to 4.30) were 2.3‐ to 8.7‐fold higher in Pgp‐deficient mice compared with wild‐type mice. In contrast, hydroxyzine was not a Pgp substrate in vitro (B → A/A → B ratio = 0.86), had high passive permeability (Papp B → A + GF120918 = 296nm/s), and had brain‐to‐free plasma concentration ratios >73 in both Pgp‐deficient and wild‐type mice. These studies demonstrate that Pgp‐mediated efflux and passive permeability contribute to the low cetirizine brain concentrations in mice and that these properties account for the differences in the sedation side‐effect profiles of cetirizine and hydroxyzine. © 2003 Wiley‐Liss, Inc. and the American Pharmacists Association J Pharm Sci 92:2082–2089, 2003