Emergency contraception, also known as the morning-after pill, is a type of birth control that can prevent a pregnancy from occurring if taken within 72 hours after sexual activity. Historically, access to emergency contraception was controlled through prescriptions by physicians. Between 1998 and 2006 eight states passed laws permitting behind the counter (BTC) access to EC; on August 24, 2006, the FDA approved sales of emergency contraception BTC through pharmacies to individuals over the age of 18 nationwide. While the BTC designation for emergency contraception indeed expanded access to individuals over the age of 18, some barriers persist for Medicaid enrollees because not all states reimburse nonprescription coverage of emergency contraception through Medicaid. Thus, the cost of emergency contraception available to low-income women on Medicaid differs by both their age and by state. The existing limitations to access by Medicaid status could mitigate the effect on unintended pregnancy. This, perhaps unintended, barrier is important since approximately 50 percent of births are Medicaid births. To the extent that BTC access can reduce unintended pregnancy, this will reduce social costs of unintended pregnancies and costs to Medicaid. This could also translate into decreased costs for these mothers, their families, their infants and later children, and public assistance participation both at birth and into the future. Using data from the Pregnancy Risk Assessment and Monitoring System (PRAMS), this study aims to understand the heterogeneous effects of changes in access to emergency contraception on the likelihood that a birth resulted from an “unintended" pregnancy. Increased access to emergency contraception should reduce unintended pregnancies by making contraceptive access easier and faster. Because the state and federal policy changes have differential applicability by subgroup, we estimate these effects for all women, women by age, and women by insurance status. We use difference-in-difference as well as difference-in-difference-in-difference methods to estimate the treatment effects.