Previously we have shown that all SELH/Bc mouse embryos close their anterior neural tubes by an abnormal mechanism and that 10-20% of SELH/Bc embryos are exencephalic. The purposes of these studies were (1) to observe the effects of retinoic acid on the frequency of exencephaly in SELH/Bc embryos; (2) to compare the SELH/Bc response with those of normal strains and of other neural tube mutants; and (3) to compare, between SELH/Bc and a normal strain (SWV/Bc), the effects of retinoic acid on morphology of the closing anterior neural tube. SELH/Bc was more liable to retinoic acid-induced exencephaly than were normal strains. After maternal treatment with 5 mg/kg retinoic acid on day 8.5 of gestation, 53% of SELH/Bc embryos had exencephaly, compared with 22% in ICR/Bc and 14% in SWV/Bc. When these results were transformed according to the assumptions of the developmental threshold model, the effects of genotype and retinoic acid appeared to be additive. Similar treatment on day 9 or 10 of gestation had little or no effect on the frequency of exencephaly in SELH/Bc mice. These results are similar to the reported responses of the curly-tail and Splotch mutants, where frequencies of spina bifida but not exencephaly were decreased. This pattern suggests that studies of effects of periconceptional vitamin treatment on risk of human neural tube defects should consider anencephaly and spina bifida separately. The study comparing the morphology of anterior neural tube closure in SELH/Bc and normal SWV/Bc embryos showed that retinoic acid delays the elevation of the mesencephalic neural folds. This results in a "stalling" of many embryos in the first steps of neural tube closure, with their neural folds remaining convex and splayed wide apart. The delay in fold elevation was superimposed on the different closure patterns of the two strains. The overall conclusion is that there is no nonadditive interaction in the parameters studied between retinoic acid treatment and the SELH/Bc genotype.