Purpose: To quantify maternal use of atypical antipsychotics, typical antipsychotics, anticonvulsants, and lithium during pregnancy. Methods: Tennessee birth and death records were linked to Tennessee Medicaid data to conduct a retrospective cohort study of 296817 women enrolled in Tennessee Medicaid throughout pregnancy who had a live birth or fetal death from 1985 to 2005. Results: During the study time period, the adjusted rate of use of any study medication during pregnancy increased from nearly 14 to 31 per 1000 pregnancies (β=0.08, 95% CI=0.07, 0.09). Significant increases were reported in use of anticonvulsants alone among mothers with pain and other psychiatric disorders, atypical antipsychotics alone among mothers with bipolar disorders, schizophrenia, unipolar depressive disorders, and other psychiatric disorders, and more than one studied medication for mothers with epilepsy, pain disorders, bipolar disorders, unipolar depressive disorders, and other psychiatric disorders. Significant decreases were reported in use of lithium alone and typical antipsychotics alone for all clinically meaningful diagnosis groups. Conclusions: There was a substantial increase in use of atypical antipsychotics alone, anticonvulsants alone, and medications from multiple studied categories among Tennessee Medicaid-insured pregnant women during the study period. Further examination of the maternal and fetal consequences of exposure to these medications during pregnancy is warranted. © 2012 John Wiley & Sons, Ltd.