Objective: To evaluate a graduated embryo scoring system for predicting blastocyst formation from cleavage-stage embryos. Design: Ongoing, prospective analysis. Materials and Methods: Two hundred and sixty-eight fertilized embryos were derived from 22 patients following IVF/ICSI and were cultured individually in P1 until day 3 and then in Blastocyst Medium until day 5. Weighted scores out of 50 were awarded at each of four evaluations based on the presence of critical developmental milestones. The first evaluation at 16–18 hours post ICSI rated cytoplasmic halo and vacuoles, pronuclear size and juxtaposition, nucleolar alignment, and polar body apposition and fragmentation. At 25–27 hours the embryos were evaluated for dissolution of the pronuclear membrane, blastomere cleavage, symmetry and the degree of fragmentation. At 40–43 hours, the embryos were evaluated for blastomere number, symmetry and percent fragmentation. A final evaluation of blastomere number and morphology at 64–67 hours generated a total GES score out of a possible 200 points. This GES score was correlated with blastocyst development. Results: A total GES score of 150–200 on day 3 was achieved by 35% (93/268) of embryos, 100–149 by 45% (121/268), and <100 by 20% (54/268). Seventy-five percent (70/93) of embryos scoring 150–200 developed to blastocyst, compared to 45% (55/121) of embryos scoring 100–149 (p<0.0001), and to 13% (7/54) scoring <100 (p<0.00001). The difference in blastocyst development between embryos scoring 100–149 and those scoring <100 was also significant (p<0.0001). Fifty-seven embryos were transferred into the 22 patients on day 5 and an additional 20 were frozen, yielding 77 (29%) “usable” embryos. Based on traditional morphology characteristics 184 would have been deemed transferable/freezable on day 3. The mean GES value for “usable” embryos was 146, compared to 129 for those not transferred or frozen. Conclusions: Graduated embryo scoring based on early developmental milestones is an accurate predictor of blastocyst development from cleavage-stage embryos. The ability to identify those embryos on day 3 which are likely to develop to blastocysts may decrease laboratory culture expense, allow earlier cryopreservation of supernumeray embryos, and could be useful in situations where the ability to culture more embryos than are intended for transfer is limited. Our early experience also indicates that there was no effect on the rate blastocyst development from individual embryo culture or from the repeated removal of the embryos from the incubator. We hope to better define those characteristics that are critical to blastocyst development, so that “usable” embryos could be identified even earlier. Data correlating GES with IVF outcome will also be presented.