This study examines public school characteristics, and public and private school market characteristics, associated with participation among elementary-aged students in a means-tested school voucher program in Florida. Participants are more likely than eligible nonparticipants to come from disadvantaged public schools on multiple dimensions. On average, participants' public schools have lower aggregate student achievement, and higher rates of disciplinary incidents. Participants' schools receive less positive ratings on various measures from principals and teachers. Participants face more competitive private school markets, and less competitive public school markets, than do nonparticipants. When these factors are considered together, the mean achievement and disciplinary incident rate of students' own public schools, as well as public and private school market variables, independently predict voucher use.