The urinary excretion time courses of pyrene-1,6-dione (P16D), pyrene-1,8-dione (P18D) and 1-hydroxypyrene (1-OHP) were compared in Sprague–Dawley and Wistar rats. Groups of five male rats, of about 200 g of body weight, were injected intravenously with 0.05, 0.5, 5 and 50 µmol pyrene kg−1 of body weight. Urine was collected at 2, 4, 6, 8, 10, 12, 18, 24, 30, 42 and 48 h post-dosing. Pyrene metabolites were measured by high-performance liquid chromatography (HPLC)/fluorescence after enzymatic hydrolysis of the glucurono- and sulfo-conjugates, extraction on Sep-Pak C18 cartridges and, for the analysis of dione metabolites, derivatization to stable diacetoxypyrene molecules. Over the 48-h sampling period, on average 17.4–25.6% of the injected pyrene was excreted overall as P16D, 6.4–8.8% as P18D and 0.6–0.8% as 1-OHP in the urine of Sprague–Dawley rats. By comparison, on average 10.3–14.7% of the intravenous pyrene dose was recovered as P16D, 4.8–6.4% as P18D and 0.3–0.4% as 1-OHP in the urine of Wistar rats. In both strains of rats there was no clear effect of the dose on the 0–48-h cumulative urinary excretion of P18D and 1-OHP over the entire dose range, while the percentage of dose recovered overall as P16D in urine at the highest dose (50 µmol kg−1) was statistically lower than at the other doses. The 0–48-h cumulative percentage of pyrene dose excreted as metabolites in the urine of Sprague–Dawley rats was also significantly higher than in Wistar rats (p<0.01) exposed under identical conditions. As for the urinary excretion-time courses of the different metabolites, for a given dose and strain of rats, excretion curves of P16D, P18D and 1-OHP generally evolved in parallel. There was also no clear effect of the dose on the excretion rate, thus half-life, of pyrene metabolites, except for P16D in Sprague–Dawley rats at the highest dose where elimination tended to be slower compared with the other doses (p<0.01). The average first-order elimination half-life of P16D, P18D and 1-OHP was 4.0, 5.7 and 4.1 h, respectively, in Sprague–Dawley rats, and 5.1, 6.1 and 5.1 h, respectively, in Wistar rats (all doses combined but excluding the highest dose for P16D). This study showed the relative importance of metabolic pathways leading to diones compared with 1-OHP. These dioxygenated metabolites appear to be interesting biomarkers of pyrene exposure at environmentally and occupationally relevant doses. Their adequacy as biomarkers of human exposure has yet to be confirmed.