A retrospective cohort study was made of the racing performance of trotters which had been subjected routinely to radiography before they started training and racing. Sixty-one per cent (148) of the 243 horses, foaled in 3 consecutive years, had one or more abnormal findings categorised into 5 relevant groups based on radiography, of which osteochondrosis (OCD) was the most specific diagnosis. Parameters used to reflect racing performance were: proportion of horses starting in races, number of starts per year, earnings per year, earnings per start and racing longevity. No significant association between the presence or type of radiological abnormalities and the subsequent performance and longevity could be found. Horses with multiple lesions, however, had a tendency to lower earnings and poorer survival than horses with single lesions.