Understanding the factors that affect per cover pregnancy rates on commercial breeding farms is important for efficient operation and improving breeding management. Breeding shed records for 9 seasons (2009–2017) from a commercial Thoroughbred breeding farm were used to evaluate factors influencing reproductive success. Observations of individual covers (n = 10,156) were recorded by the manager and trained staff and included information on mating demographics; cover and semen characteristics; and mare conformation and behavior. Incomplete records were removed from the data set providing 8,843 observations. The effect of each factor on pregnancy rate was analyzed using a generalized linear model and binomial distribution. Per cover pregnancy rate (PR) was different (P < 0.001) between stallions (n = 20) and ranged from 53 to 70% (mean average = 63%). PR was lower (P < 0.001) in barren mares compared with foaling and maiden mares; July covers (P < 0.001) compared with all other months; and when dismount samples contained urine (P < 0.001). Mares scoring 0 or 1 (on a 0–4 tease scale) were less likely (P < 0.05) to conceive than mares showing stronger estrous behaviors when teased before mating with LS means of 40.4, 54.4, 62.2, 62.6 and 65.9 % PR for tease scoresof 0, 1, 2, 3 and 4 respectively. There was a trend (P = 0.08) for lower PR in mares requiring tranquilization (57.1%) compared with non-tranquilized mares (62%) supporting the importance of estrous behavior for successful mating. Insemination post live cover was performed 15% of the time due to either a large dismount sample or per owner request, however there was no difference in PR. Pregnancy rate was higher (P < 0.05) in mares with vulva sutures (63.5%) or having acceptable vulvar conformation without sutures (64.1 %) compared with poor vulva conformation without sutures (59.3%). Mating quality assessed by a subjective cover grade (A, B+, B, B-, C) positively influenced PR (P < 0.001) while other factors including: number of mounts, number of semen pulses, time of day and presence of RBC and WBC in the dismount sample had no effect on PR, although there was a trend (P < 0.08) for lower PR when WBCs were noted in the dismount samples. Maintaining detailed breeding shed records can provide managers with useful information which could be used in making management decisions.