In Experiment (Exp.) 1 the survival of 8- to 16-celled cow embryos (early morulae) and late morulae after cooling to 0.degree. C was compared. Only 1/19 early morulae developed (in rabbit oviducts) to the blastocyst stage after being cooled to 0.degree. C, whereas 8/11 developed after being stored at 20.degree. C (.chi.2 = 12.06, P < 0.001). In contrast, 7/9 late morulae which had been cooled to 0.degree. C developed to blastocysts as did all 3 late morulae which had been left at 20.degree. C. In Exp. 2, a comparison was made of the ability to develop to calves of late morulae which had and had not been cooled to 0.degree. C. Twelve late morulae were transferred in pairs to 6 non-pregnant recipients; 3 heifers subsequently gave birth to 4 normal calves, 2 of which had developed from control morulae and 2 from cooled morulae. In Exp. 3, 42 late morulae were frozen and thawed by a method developed for the low temperature preservation of mouse blastocysts (Wilmut, 1972). The concentration of dimethylsulfoxide (DMSO) was increased to 1.5 M in 3 equal steps, at 10-min intervals at 20.degree. C. The DMSO was removed in either 3 or 6 steps before the embryos were transferred to ligated rabbit oviducts; in some instances the embryos were transferred directly after thawing. Twenty-six embryos were recovered 1 day later and 2 had developed to normal blastocysts. Probably, most early morulae of cows are killed by being cooled relatively rapidly to 0.degree. C, and there probably is variation between species and variation between the stages of development of the embryos within a species in their sensitivity to cooling to 0.degree. C. These effects may be critical in the development of satisfactory methods for the storage of embryos at very low temperatures.