Rat embryos were explanted either on day 8, 3 PM (primitive streak stage) or on day 9, 8 AM (presomite stage) and divided randomly into two groups (plug day = day 0). In one group, Reichert's membrane was removed starting from the side opposite to the ectoplacental cone which was left intact (standard method). In the second group, Reichert's membrane was removed starting from the same side as the ectoplacental cone after the latter structure was excised and the proamniotic cavity (day 8) or the ectoplacental cavity (day 9) destroyed (new method). Embryos were then cultured and examined at 10 AM on day 11. All embryos developed normally in size and general morphology. However, there was a high incidence of inverted heart, allantoic placenta and tail in all groups, except for day 9, new method group; this was the only group in which the ectoplacental cavity did not expand abnormally during culture. The new method, which is quick and easy to perform even when Reichert's membrane does not protrude well, lessens the chances of damaging embryonic ectoderm during the removal of Reichert's membrane. Furthermore, the absence of ectoplacental cone decreases embryo aggregation during culture. We hypothesize that the abnormally expanded ectoplacental cavity generates abnormal physical forces within the egg cylinder that disrupt the control of body asymmetry.