Intranasal delivery of flurazepam, midazolam, and triazolam was studied in a dog model as a possible alternate route of drug administration for treatment of insomnia. Four beagles received each hypnotic by both intranasal and oral routes on two separate occasions. Plasma concentrations for each hypnotic after dosing were measured by electron-capture gas-liquid chromatography. The mean intranasal absorption rates (tmax) of flurazepam, midazolam, and triazolam were 1.7, 2.0, and 2.6 times faster, respectively, compared with oral dosing. The mean dose-normalized peak concentrations (Cmax) after intranasal delivery were 16.4, 2.9, and 3.4 times higher, respectively, versus oral administration. The mean dose-normalized AUCs estimated for these compounds after nasal administration were 2.4-, 2.5-, and at least 2-fold larger than after oral administration for midazolam, triazolam, and flurazepam, respectively. If these observations can be extrapolated to humans, the faster absorption achieved by the intranasal route would appear to benefit insomniacs characterized by difficulty in falling asleep because of an anticipated faster sedative effect onset. The higher peak concentrations and larger amounts absorbed in the case of intranasal midazolam and triazolam delivery may lead to dose reduction.