SummaryBackgroundOsteochondrosis is considered multifactorial in origin, with factors such as nutrition, conformation, body size, trauma and genetics thought to contribute to its pathogenesis. Few studies have investigated the effects of genetic variability of osteochondrosis in Thoroughbreds.ObjectivesTo describe the prevalence and genetic variability of a subset of osteochondrosis lesions in a group of Thoroughbred yearlings.Study designRetrospective cohort study.MethodsRadiographs of 1962 Thoroughbred yearlings were retrieved from clinical records obtained between 2005 and 2013. Pedigree information was obtained from the Australian Stud Book. Osteochondrosis lesions were documented in selected joints and estimates of heritability were obtained by fitting linear mixed models in ASREML software.ResultsThe overall prevalence of osteochondrosis was 23%. Osteochondrosis was identified in 10% of stifle joints, 6% of hock joints and 8% of fetlock joints. The heritability estimates ranged from 0 to 0.21. The largest estimates were 0.10, 0.14, 0.16 and 0.21 for lesions of the distal intermediate ridge of the tibia, dorso‐proximal proximal phalanx (P1), any stifle osteochondrosis, and lesions of the lateral trochlear ridge of the distal femur, respectively. Although calculated heritability estimates had high standard errors, meta‐analyses combining the present results with published estimates were significant at 0.10, 0.17, 0.15 and 0.20 for stifle, tarsal, fetlock and these joints combined, respectively. In addition, there was a permanent environment attributable to the dam effect.Main limitationsInclusion criteria were based on radiographic findings in specific joints at a specific age range in Thoroughbreds.ConclusionsThe present results indicate that only a proportion of osteochondrosis in Thoroughbreds is heritable. The permanent environment effects of the dam were observed to have effects on some categories of osteochondrosis.