BACKGROUND: Planning a home birth does not necessarily mean that the birth will take place successfully at home. The object of this study was to describe reasons and risk factors for transfer to hospital during or shortly after a planned home birth. METHODS: A nationwide study including all women who had given birth at home in Sweden between January 1, 1992, and July 31, 2005. A total of 735 women had given birth to 1,038 children. One questionnaire for each planned home birth was sent to the women. Of the 1,038 questionnaires, 1,025 were returned. Reasons for transfer and obstetric, socioeconomic, and care-related risk factors for being transferred were measured using logistic regression. RESULTS: Women were transferred in 12.5 percent of the planned home births. Transfers were more common among primiparas compared with multiparas (relative risk [RR] 2.5; 95% CI 1.8-3.5). Failure to progress and unavailability of the chosen midwife at the onset of labor were the reasons for 46 and 14 percent of transfers, respectively. For primiparas, the risk was four times greater if a midwife other than the one who carried out the prenatal checkups assisted at the birth (RR 4.4; 95% CI 2.1-9.5). A pregnancy exceeding 42 weeks increased the risk of transfer for both primiparas (RR 3.0; 95% CI 1.1-9.4) and multiparas (RR 3.4; 95% CI 1.3-9.0). CONCLUSIONS: The most common reasons for transfer to hospital during or shortly after delivery were failure to progress followed by the midwife's unavailability at the onset of labor. Primiparas whose midwife for checkups during pregnancy was different from the one who assisted at the home birth were at increased risk of being transferred.