The aim of this study was to characterize the incidence and timing of pregnancy loss from service event (timed artificial insemination or timed embryo transfer) to parturition. Lactating Holstein-Friesian cows were randomly assigned to receive either artificial insemination (AI, n = 243) or embryo transfer (ET, n = 863) with a fresh or frozen in vitro-produced blastocyst derived from oocytes collected from the ovaries of elite dairy donors (n = 14 Holstein Friesian and n = 8 Jersey) and elite beef donors (n = 21 Angus) using transvaginal ovum pick-up. In addition, oocytes were collected from the ovaries of beef heifers of known pedigree following slaughter at a commercial abattoir (n = 119). Blood samples were collected on d 7 from all cows to determine progesterone concentration and from subsets of cows on d 18 (n = 524) and d 25 (n = 378) to determine mRNA abundance of interferon-stimulated gene-15 and pregnancy-specific protein B concentration, respectively, to provide an early pregnancy diagnosis. Transrectal ultrasonography was conducted to determine pregnancy status on d 32, d 62 and d 125 after synchronized ovulation. Parturition date was recorded for all cows that reached a term delivery. The predicted probability of pregnancy (%) varied at each time point (d7, 18, 25, 32, 62, 125, parturition) depending on treatment (AI: 77.0, 60.2, 52.3, 48.8, 47.0, 44.6, 44.0; fresh ET: 100.0, 69.5, 60.3, 56.1, 48.4, 46.8, 45.5; frozen ET: 100.0, 61.7, 52.2, 41.6, 32.9, 31.8, 30.2). Irrespective of treatment, the largest proportion of pregnancy loss occurred in the period from service event (AI on d 0 or ET on d 7) to d 18, with minimal loss occurring between d 62 and parturition (AI: 1.8%, fresh ET: 1.9%, frozen ET: 3.5%). Treatment differences in the predicted probability of pregnancy per service event were detected between fresh ET vs frozen ET on d 32 and both AI and fresh ET vs frozen ET on d 62, 125 and at parturition. There was greater probability of pregnancy loss between d 32 and 62 following ET (Fresh: 11.3%, Frozen: 18.0%) than AI (4.0%). The percentage of cows that calved following transfer of a fresh embryo (45.5%) was similar to AI (44.0%), but less when a frozen embryo was transferred (30.2%). In conclusion, AI and fresh ET led to a greater probability of a cow becoming pregnant and maintaining the pregnancy to term than frozen ET. Cows that were still pregnant on d 62 had a very strong likelihood of maintaining the pregnancy to full-term parturition, irrespective of treatment. Further work is required to improve the likelihood of pregnancy establishment and reduce embryonic and fetal mortality following transfer of a cryopreserved in vitro-produced embryo.