The objective of this study was to evaluate the use of a second flush for bovine embryo recovery. A total of 319 clinically healthy Holstein cows (247 lactating, 53 dry, 19 nulliparous) with an average age of 5.5 ± 2.5 years were used for this experiment. Superovulation was performed according to a modified method of Baracaldo et al. (2000). On Day 0 (beginning of the experiment), each cow received 3 mg of estradiol-17β intramuscularly (i.m.) and a progesterone-releasing vaginal insert (1.9 g of progesterone, CIDR, Pfizer Animal Health, Kirkland, QC, Canada) at random stages of the estrous cycle. From Day 4 evening to Day 8 evening, the cows received a total of 380 mg of NIH-FSH-P1 (FolltropinV, Bioniche Animal Health) administered im through 9 injections of decreasing dose (from 70 to 20 mg) at 12-h intervals. On Day 7, the cows received 2 injections consisting of 500 μg of cloprostenol (prostaglandin F_2α analogue Estrumate, Schering-Plough, Pointe-Claire, QC, Canada) given approximately 12 h apart and vaginal inserts were removed 12 h after the last injection. Artificial insemination was performed on Day 10 after treatment with 100 μg, GnRH im (Cystorelin, Merial Canada Inc, Baie d’Urfe, QC, Canada). Embryos were flushed from the uterus of donor cows 6 days after AI. The method consisted of using simultaneously 1 catheter (18Fr Silicone 2-way, Bioniche Animal Health) per uterine horn. Catheters were maintained in place to perform 2 flushes 1 h apart. A total of 1000 mL of flushing media (Complete flush, Bioniche Animal Health) were used, 700 mL and 300 mL for the first and the second flush, respectively. Embryos were assessed for viability immediately after collection using IETS classification. Data were analyzed using the SAS MIXED procedure (SAS Institue, Cary, NC). The mean (±SD) number of embryos collected at the first flushing was 5.87 ± 5.1, 0.92 ± 2.2 and 2.9 ± 4.4 for transferable, degenerate and unfertilized oocytes, respectively. The second flushing yielded 2.32 ± 2.6 transferable embryos, 0.28 ± 0.83 dead embryos and 1.2 ± 2.2 unfertilized oocytes. There was no significant effect of age, day in milk, or stage of lactation on transferable or degenerate embryos or nonfertilized oocytes in each flushing. The embryo recovery method used in this experiment could be used to recover more transferable embryos.