Embryo transfer in horses started in the 1970′s and has evolved since then in several technical and biological areas that may influence the overall outcome. The transfer method evolved from surgical to non-surgical (NSET), with pregnancy rates for NSET related closely to operator skills. Recently, high pregnancy rates have been reported by non-specialized operators using the cervical forceps method (Cuervo Arango et al., Veterinary Record. 2018,183;10,1-5). The aim of this retrospective study was to compare pregnancy and early embryonic loss rates between conventional and cervical forceps transfer techniques (Wilsher and Allen, Equine Vet. Educ. 2004. 16; 1, 39-44), operated by two veterinarians with different degrees of experience. During the last 12 breeding seasons in a commercial equine embryo transfer program in Chile, 345 embryos were transferred by two operators with low / moderate (n=86) or high experience (n=259), using a conventional manual technique and 228 embryos were transferred by the same two operators (n=81) and (n=147) respectively, using a bovine cervix forceps/Polansky speculum technique. Embryo flushes were performed at 8 or 9 days after ovulation on 102 donor mares aged 4 to 24 years old and transferred to recipient mares (n=387) aged 3 to 12 years old. Ovulation synchrony ranged from -1 to +4 days when comparing the recipient to the donor. No medications were used during embryo transfer and procedures were done at the samefacilities. Data were analyzed by a binary logistic regression model in which pregnancy and embryo loss were treated as dependent variables, and the experience of the operator and the ET technique as independent variables. Pregnancy rates at first diagnosis (12 to 14 days post ovulation) and embryo loss rate until 60 days of gestation were compared between and within operator groups. Overall pregnancy rates were higher (P<0.001) by the cervical forceps / speculum embryo transfer technique (195/228; 85%) than with the conventional technique (236/345; 65%) in both groups combined. Early embryonic loss rate did not differ (P>0.1) between the groups (6.0 vs. 4.6%, respectively). For the conventional technique, there was a significant operator effect (P=0.015) on the pregnancy rates at the first diagnosis (57 vs. 72%); no operator effect (P>0.1) was observed when using cervical forceps (84 vs. 86 %). These results agree with the report of Cuervo-Arango et al. (2018). We also confirmed that simple bovine cervical forceps work as well as the equine specific forceps reported by Wilsher and Allen (2004). In conclusion, the cervical forceps technique yielded higher pregnancy rates, with no effect of the operator's experience compared to the conventional technique in which the skill and experience of the operator is crucial and needs time and losses to achieve.