BACKGROUND This study evaluates the time-of-day effect on midazolam and 1-OH midazolam pharmacokinetics, and on the sedative pharmacodynamic response in rabbits. Also, circadian fluctuations in rabbits' vital signs, such as the blood pressure, heart rate and body temperature were examined. The water intake was measured in order to confirm the presence of the animals' diurnal activity. The secondary aim involved the comparison of two methods of data analysis: a noncompartmental and a population modeling approach. METHODS Twelve rabbits were sedated with intravenous midazolam 0.35 mg/kg at four local times: 09.00, 14.00, 18.00 and 22.00 h. Each rabbit served as its own control by being given a single infusion at the four different times of the day on four separate occasions. The values of the monitored physiological parameters were recorded during the experiment and arterial blood samples were collected for midazolam assay. The pedal withdrawal reflex was used as the measurement of the sedation response. Two and one compartmental models were successfully used to describe midazolam and 1-OH midazolam pharmacokinetics. The categorical pharmacodynamic data were described with a logistic model. CONCLUSIONS We did not find any time-of-day effects for the pharmacokinetic and pharmacodynamics parameters of midazolam. For 1-OH midazolam, statistically significant time-of-day differences in the apparent volume of distribution and clearance were noticed. They corresponded well with the rabbits' water intake. The noncompartmental and model-based parameters were essentially similar. However, more information can be obtained from the population model and this method should be preferred in chronopharmacokinetic and chronopharmacodynamic studies.