Pharmacokinetics is the study of the time course of a drug and its met
abolites following its introduction into the body. Population pharmaco
kinetic studies are becoming increasingly important as an aid to drug
development. The data from such studies typically consist of dose hist
ories, drug concentrations with associated sampling times, and often c
ovariate measurements such as the age and weight of each subject. Thes
e studies aim to provide an understanding of the pharmacokinetics of t
he drug in question and so lead to an informed choice of dosage regime
n. Such an understanding includes determining those covariates that ar
e important predictors of fundamental pharmacokinetic parameters, such
as clearance, defined as the volume of plasma cleared of drug in a un
it of time. Determining those subpopulations (e.g., the elderly) with
altered kinetics has implications for the choice of an appropriate dos
age regimens, because predictive concentration profiles arising from a
particular regimen in different populations may be very different. In
this article a general Bayesian hierarchical model is described. Phar
macokinetic models relating concentration to time are generally nonlin
ear, and the data are often sparse and/or noisy. The number of individ
uals on whom data have been collected is often large, and so the dimen
sionality of the parameter space is large. Consequently, estimation, f
rom a Bayesian or a classical perspective, is not straightforward. In
this article the Hastings-Metropolis algorithm is used for learning ab
out the posterior distribution. An analysis of concentration data coll
ected after the administration of the antiarrhythmic drug quinidine is
presented. The data consist of 361 measurements on a total of 136 pat
ients. Nine covariates are also available for each individual. These c
ovariates are a mixture of discrete and continuous measurements. Some
of the covariates are constant within an individual during the course
of the study, whereas others change. A covariate model is constructed,
and the sensitivity of the inferences to distributional assumptions i
s examined. The importance of assessing the appropriateness of modelin
g assumptions is emphasized and extensive model checking is carried ou
t for the quinidine data using graphical diagnostics.