Using data from the San Diego Birth Center Study that enrolled underserved women between 1994 and 1996, we examined demographic, sociobehavioral, and medical predictors of hospital transfer in a group of women who intended to deliver at a freestanding birth center. Of the 1808 women, 34.6% transferred to the hospital antenatally and 19.6% transferred during labor, while 45.7% delivered at the birth center. Compared with multiparous women who had never had a cesarean and never had a previous hospital delivery, nulliparous women were 2.0 times more likely (95% confidence interval [CI], 1.4–2.7), multiparous women with a previous cesarean were 2.6 times more likely (95% CI, 1.7–3.8), and women without a previous cesarean but who had a previous hospital delivery were 2.1 times more likely (95% CI, 1.5–3.0) to transfer after adjusting for other predictors of transfer. Nulliparity, cesarean history and having a previous hospital delivery were among the strongest predictors of a hospital transfer even after adjusting for demographic, sociobehavioral, and other medical conditions. Understanding predictors of transfer may assist practitioners, patients, and policy makers in considering the appropriateness of individuals for birth center delivery or to target further education to reduce nonmedical transfers.