Pregnant A/WySn mice, 20 to 30% of whose offspring have spontaneous cleft lip, were treated with thyroxine. Following treatment, cleft lip and normal embryos died, but cleft lip embryos died at a higher rate. The increased liability of cleft lip embryos to thyroxine-induced death was considered as a possible experimental route to identify the basic genetic defect that causes cleft lip. A time-response study indicated that cleft lip embryos responded more than normals following treatment on any of days 7 to 12 of gestation, that there is no sharply defined critical period, and that normal and cleft lip embryos do not differ in time of maximum sensitivity. A dose-response study showed linear responses of normal and cleft lip embryos on a probit-log dose scale, with a common slope and LD50's of 1.9 and 1.3 mg respectively. These dose-response properties indicate that normal and cleft lip embryos are probably killed by the same mechanism, but differ in dosage tolerance. That is, they differ quantitatively, not qualitatively. Thyroxine did not significantly change the cleft lip frequency, and the difference between normal and cleft lip embryos that leads to cleft lip itself is therefore not in the same pathway as that which leads to thyroxine-induced death. A hypothetical example of the defect basic to both pathways is presented.