This study was performed in order to determine optimum flushing solution using the direct embryo collection (DEC). Donors, at random stages of the estrous cycle, received a CIDR. 7 days later, 200 mg FSH was treated with 40, 30, 20, 10 mg FSH levels in declining doses twice daily by intramuscular injection for 4 days. On the 3$^{rd}$ day administration of FSH, 25 mg $PGF_2{alpha}$ was administered and CIDR was withdrawn. After FSH injections were complete, donors were artificially inseminated twice at 12 hr intervals. The donor cattle received 250 ${mu}g$ GnRH at time of 1$^{st}$ insemination and embryos were recovered 8 days after the 1$^{st}$ insemination. Embryo collection from superovulated donors were performed to flushing by DEC and conventional method. As a results, the average number of recovered embryos were significantly higher as 19.1${pm}$1.40 with DEC method than 12.0${pm}$0.44 with conventional embryo collection method, respectively (p<0.05). Also, The average number of transferable embryos were significantly higher (p<0.05) as 15.8${pm}$1.72 with DEC method than 6.9${pm}$0.35 from conventional embryo recovery procedures. Meanwhile, number of recovered embryos and number of recovered transferable embryos following the number of flushing times until 6${dr}$ flushing were significantly higher as 8.6${pm}$0.53 and 8.6${pm}$0.53 from 2$^{nd}$ flushing time than other groups (p<0.05). No. of Ear. B stage embryos were significantly higher as 3.9${pm}$0.90 and 3.9${pm}$0.90 with 2$^{nd}$ flushing time in total collected embryos and transferable embryos (p<0.05). Com M stage embryos were significantly higher as 3.7${pm}$1.00 in 2$^{nd}$ flushing time and as 2.2${pm}$0.76 in 3$^{rd}$ flushing time for recovered embryos (p<0.05). In transferable embryos, Com. M stage embryos were significantly higher (p<0.05) as 3.7${pm}$1.00 in 2$^{nd}$ flushing time and as 2.2${pm}$0.76 in 34$^{dr}$ flushing time, also. No. of degradation embryos was significantly higher as 2.2${pm}$0.72 in 5${rd}$ flushing time, On the other hand, degradation embryos was not observed in transferable embryos (p<0.05). In conclusion, these results suggest that DEC method should effective methods for production of in vivo embryos using less flushing solution following perform until 4$^{rd}$ flushing time than conventional embryo collecting method. Also, it might be effectively collection of transferable embryos following more less procedure times compared to conventional embryo recovery methods.