Angular and flexural limb deformities (LD) are important causes of early-life morbidity and mortality in Thoroughbred foals. There is believed to be an important congenital component however, the exact etiologies remain unknown. We hypothesized that maternal- and pregnancy-level factors, particularly those with potential to influence in-utero growth and development, could play an important role. The aims of this study were to estimate the incidence of early-life LD in foals and to investigate associations with mare- and pregnancy-level factors. A Thoroughbred birth cohort was established on seven farms across the United Kingdom and Ireland and details of veterinary interventions for LD in foals in the first 6-months of life prospectively recorded. Where available, details of dams’ signalment, breeding history and reproductive history in the breeding season(s) of interest were retrospectively retrieved from stud and veterinary records. Analyses were carried out at the pregnancy resulting in the birth of a live foal-level. Incidence risk of LD was estimated as number of new cases requiring veterinary intervention at least once within the first 6-months of life, and associations with mare and pregnancy factors were assessed using mixed effects, multivariable logistic regression modelling. Data were available for 275 pregnancies, in 235 mares sired by 89 stallions over two breeding seasons (2018 and 2019). Records allowed for the evaluation of nine factors as fixed effects (the mares’residing stud farm, age, status and number of previous live foals, month of covering, breeding season, fetal sex, gestation length and whether dystocia was reported). Pregnancies resulted in the birth of 272 live foals (n=3 died during parturition), 21% of which (n=57/272, 95% confidence interval 16-26) required veterinary intervention at least once for LD in the first 6-months of life. Seventy-five percent (n=43/57, 95%CI 63-87) were described as flexural and 25% (n=14/57, 95%CI 15-37) as angular deformities; median age at veterinary intervention was 9 days (inter-quartile range 1-19). Only gestation length was retained in the final model; model fit was not improved by inclusion of farm, mare or stallion as a random effect (likelihood ratio tests P=0.10, P=0.36 and P=0.92, respectively). Odds of LD decreased by 4% per day increase in gestation length between 314-381 days (OR 0.96, 95%CI 0.93-0.99, P=0.01). The main limitations of this work were the retrospective nature of the mare data, which were not recorded specifically for this study, and the use of veterinary records to identify disease, which may underestimate the true incidence of LD in the population. In conclusion, longer gestation length appears to reduce the odds of early-life LD in Thoroughbred foals. Further work is required to elucidate the biological mechanisms behind this association.