Every program of assisted reproduction strives to maximize pregnancy outcomes from in vitro fertilization and selecting an embryo culture medium, or medium pair, consistent with high success rates is key to this process. The common approach is to replace an existing medium with a new one of interest in the overall culture system and then perform enough cycles of IVF to see if a difference is noted both in laboratory measures of embryo quality and in pregnancy. This approach may allow a laboratory to select one medium over another but the outcomes are only relevant to that program, given that there are well over 200 other variables that may influence the results in an IVF cycle. A study design that will allow for a more global application of IVF results, ones due to culture medium composition as the single variable, is suggested. To perform a study of this design, the center must have a patient caseload appropriate to meet study entrance criteria, success rates high enough to reveal a difference if one exists and a strong program of quality assurance and control in both the laboratory and clinic. Sibling oocytes are randomized to two study arms and embryos are evaluated on day 3 for quality grades. Inter and intra-observer variability are evaluated by kappa statistics and statistical power and study size estimates are performed to bring discriminatory capability to the study. Finally, the complications associated with extending such a study to include blastocyst production on day 5 or 6 are enumerated.