We recently conducted a dose-response study of the effects of cocaine on several activity measures in the panel of BxD/Ty recombinant inbred mice. Animals were tested in an automated activity chamber over 2 days with i.p. saline on day 1 and i.p. cocaine on day 2, at one of four doses, 5, 15, 30 or 45 mg kg(-1). The monitor recorded total distance traveled, nosepokes in a holeboard, repeated movements and time spent by an individual in proximity to the centre of the apparatus. Dose-response curves for locomotor activation, i.e. the difference between cocaine and saline scores, showed that for all strains tested, scores increased 5-30 mg kg(-1). With few exceptions, locomotor activity at 45 mg kg(-1) was not significantly higher than that at 30 mg kg(-1). Repeated movement scores showed patterns similar to locomotor activity and nosepokes tended to be progressively inhibited by increasing doses of cocaine. Recombinant inbred strain mean distributions for all behaviours and at all doses exhibited continuous, rather than discrete variation, thus providing evidence of multiple-gene effects on cocaine-related behaviours. Quantitative trait loci (QTL) analysis pointed to several chromosomal locations associated with variations in cocaine-related behaviours and some are either identical or close to QTL reported by others. In separate groups of animals, densities of dopamine D1, and D2 receptors and dopamine uptake transporters were measured in the medial prefrontal cortex, caudate-putamen, nucleus accumbens and ventral midbrain. In all areas, all measures showed distributions consistent with polygenic influence and were associated with QTL. Of particular interest was our finding of a large segment on chromosome 15, which is related to dopamine receptor densities and cocaine-related behaviours.