In a typical developmental toxicity study, the endpoint of interest is the overall toxic effect of the chemical compound based on embryotoxicity and structural developmental toxicity. This overall toxicity is measured in terms of the numbers of resorptions, intrauterine deaths, and live pups having skeletal malformations. So far, the practice has been to study the embryotoxicity and structural developmental toxicity separately. We examine the use of Dirichlet-Trinomial model to handle the potential litter effect and also the joint distribution of the pups showing two types of toxic responses. In addition, we employ Weibull dose-response models for the two types of toxic responses described above. These dose-response models are useful in the risk assessment for a given dose and also in the calculation of safe dose levels. The likelihood ratio tests for testing equality of just the resorption/death rates or the equality of just the malformation rates or the equality of rates of overall toxic responses at various dosages can be carried out with the help of this model. The results obtained by using the Dirichlet-Trinomial model are compared with the results obtained by using the generalized trinomial model, studied by Louis Ryan, and also with the results obtained by using the beta-binomial model. The models are compared with an application to a real data set.