We selected 262 case-parent triads from a population-based study of orofacial clefts in Norway, and examined variants of developmental genes TGFA, TGFB3, and MSX1 in the etiology of orofacial clefts. One hundred seventy-four triads of cleft lip cases (CL+/-P) and 88 triads of cleft palate only cases (CPO) were analyzed. There was little evidence for an association of any of these genes with CL+/-P. The strongest association was a 1.7-fold risk with two copies of the TGFB3-CA variant (95% CI=0.9-3.0). Among CPO cases, there was a 3-fold risk with two copies of the TGFA TaqI A2 allele, and no increase with one copy. Assuming this to be a recessive effect, we estimated a 3.2-fold risk among babies homozygous for the variant (95% CI=1.1-9.2). Furthermore, there was strong evidence of gene-gene interaction. While there was only a weak association of the MSX1-CA variant with CPO, the risk was 9.7-fold (95% CI=2.9-32) among children homozygous for both the MSX1-CA A4 allele and the TGFA A2 allele. No association of CPO with the TGFA variant was seen among the other MSX1-CA genotypes. In conclusion, no strong associations were found between CL+/-P and variants at these three genes. There was a possible recessive effect of the TGFA TaqI variant on the risk of CPO, with a 3-fold risk among children homozygous for the variant. The effect of this TGFA genotype was even stronger among children homozygous for the MSX1-CA A4 allele, raising the possibility of interaction between these two genes.