The objective was to evaluate the effect of the first artificial insemination (AI) on the number of viable and degenerated embryos, total embryos and unfertilized oocytes, and fertilization rate in Holstein donor cows. Cows (n=22) and heifers (n=50) were synchronized by the Crestar® (Intervet International B.V., Boxmeer, Holland) protocol where 3 mg of norgestomet are implanted subcutaneously in addition to an intramuscular injection of 3 mg of norgestomet and 5 mg of estradiol valerate on Day 1 (D1). On the sixth day, superovulation was initiated with 8 decreasing injections of FSH/LH (PLUSET®; I.F. Serono, Rome, Italy), twice daily at 12 hour intervals. On the eighth day, 0.5 mg of sodium cloprostenol (Sincrosin®, Vallée S.A., Montes Claros, Brasil) was injected, and the implant was withdrawn on the ninth day. Embryos were flushed seven days after the first AI and classified according to the developmental stage and quality. Animals were allocated ramdomly to one of two treatments where in AI was at 0 and 12 hours (T0; n=27) and at 12 and 24 hours (T12; n=45) after onset of estrus. Fertilization rates were similar between treatments (T0 versus T12) and parity (heifers versus cows): 87.1 ± 4.8% versus 85.9 ± 3.2% and 87.1 ± 2.9% versus 85.9 ± 4.9%, respectively. There was no difference between treatments (T0 versus T12) and parity (heifers versus cows) for average number of viable embryos (5.3 ± 1.0 versus 5.7 ± 0.7 and 5.7 ± 0.6 versus 5.3 ± 1.0), unfertilized structures (1.0 ± 0.4 versus 1.2 ± 0.3 and 1.3 ± 0.3 versus 0.9 ± 0.5), and degenerated structures (1.1 ± 0.4 versus 1.2 ± 0.3 and 1.5 ± 0.3 versus 0.9 ± 0.4), respectively. There was an interaction (P = 0.04) between parity and treatment. Heifers in T12 had more viable embryos (7.1 ± 0.8) than those in T0 (4.2 ± 0.9). It was concluded that superovulated Holstein cows, but not heifers, can be inseminated at the onset of estrus without any deleterious effect on embryo yield or quality.