In this review an in vitro system is considered to be any developing tissue, organ, or organism other than mammalian embryo in situ. Before the various test systems that have been used or proposed are discussed, attention is given to the questions of whether all chemicals in the environment are in need of biological testing and what criteria should be used in making this judgment. Consideration is also given to selection of the appropriate level of rigorousness to be used in tests of different categories of chemicals. To aid in this, the characteristics of an optimal test are assembled and used as a standard for estimating the potential usefulness of the various in vitro systems. The systems discussed include bacteria and other unicellular organisms, somatic cells in culture, tissue culture, organ culture, intact invertebrate embryos (e.g., drosophila, sea urchins, sand dollars), intact lower vertebrate embryos (e.g., frogs, other amphibians, fish), cultured mammalian embryos, and incubating chick embryos. None of these are regarded as sufficiently validated in terms of comparisons with known teratogenic responses in pregnant mammals to warrant adoption as a reliable test at this time. Intact embryos of drosophila, sea urchins, amphibians, and fish are regarded as promising, but much research is needed to ascertain their predictive validity for mammals. The incubating chick embryo, however, possesses more of the essential features of the optimal system than any of the others. A tentative proposal using the chick is outlined, but it will require considerable further comparison with currently used procedures in pregnant mammals before its reliability can be fully evaluated.