The reversibility of quinolone-induced cartilage lesions has not been studied in detail. We treated five groups of five to seven juvenile Wistar rats (male and female; age: 5 weeks) with 2 x 600 mg ofloxacin/kg by gastric intubation on 1 day only (9:00 a.m. and 5:00 p.m.) and studied the knee joints histologically 3 days, 1, 3, 8 and 17 weeks later. In addition, joint cartilage specimens from vehicle-treated control rats (n = 21) at corresponding age were examined. Cartilage lesions such as matrix swelling, loss of proteoglycans and horizontal clefts were found in nearly all knee joints (26 of 27 joints; incidence: 96%) of the ofloxacin-treated rats. Within the observation period of 4 months the size of these lesions in knee joint cartilage did not decrease significantly. The diameter of the lesions at the time points of evaluation was 1146 +/- 535, 1713 +/- 309, 1250 +/- 585, 1406 +/- 356, and 1542 +/- 467 mu m, respectively (mean values +/-sd). Chondrocyte clusters producing glycosaminoglycans were observed 3 weeks after dosing and at later time points. They are considered to reflect the onset of repair but chondrocyte organization did not normalize during the study period, thus indicating the irreversibility of the effect under the experimental conditions. In principle, long-term joint cartilage damage has to be taken into account when the use of quinolones in children is considered. More detailed pharmacokinetic data are necessary for a reasonable risk assessment approach.