An investigation was undertaken to ascertain the potential teratogenicity of magnetic resonance imaging (MRI) fields. The C57BL/6J mouse was chosen as the experimental model with eye malformations (microphthalmia and morphologic anomalies) designated as the biological end point. This mouse strain is genetically predisposed to this type of malformation as a 10% spontaneous incidence occurs. Dams in groups of 15 were subjected to MRI imaging conditions on gestational day (Gd) 7 for 36 minutes to a spin-echo T-2-weighted scan by using a 1.5 Tesla magnetic field and a radiofrequency (RF) field of 64 MHz. One group was exposed at the magnetic isocenter while another was exposed at the entrance to the magnet lumen. There was also a sham control group. The dams were sacrificed at Gd 14. Assessment of eye abnormality was determined by, 1) a veterinary ophthalmologist, 2) a computer-based method comparing eye areas, and 3) a methodology combining both the above subjective and quantitative methods. MRI fields were found to produce malformation rates (15-37%) higher than controls (2-19% P less than or equal to .05, Kruskal-Wallis Test) for both isocenter and lumen entrance groups. The malformation rates and degree of statistical significance varied somewhat with analytical methodology and the unit of measure (right eye, left eye, or fetus). The results suggest for the first time the potential of MRI fields to produce developmental malformations in an animal model utilizing clinically realistic exposure conditions. (However, the reader is remained that the mouse strain utilized in this investigation was genetically prone to malformations).