The behaviour of animals in tests of anxiety varies between strains, even in identical tests and surroundings. To evaluate the results obtained, a more detailed knowledge of the behaviour of different rat strains is indispensable. Identically raised Fischer 344 rats and two stocks of Wistar rats were examined in two animal tests of anxiety: the X-maze and a modified open-field test following diazepam treatment (0.5-4.0 mg/kg). Harlan-Wistar rats were the least 'anxious' when the behaviour of vehicle treated controls was compared. The largest effect of the anxiolytic diazepam, however, was observed in Harlan-Fischer rats. To determine possible reasons for strain and stock differences, plasma concentrations of diazepam and metabolites and concentrations of serotonin (5-HT) in the CNS were measured. Plasma concentrations of diazepam and metabolites differed between the strains with the Harlan-Fischer rats showing higher diazepam concentrations. 5-HT levels in discrete brain regions varied with Harlan-Fischer rats having higher 5-HT concentrations. Strain differences influence the anxiety-associated behaviour of untreated animals and the effect of anxiolytics.