This study investigates the relationships between individual learner characteristics and gain in writing performance as measured by a standardized writing assessment in a sample of community college academic English as a Second Language (ESL) students in the United States. The ethnically and linguistically diverse convenience sample included 76 students (43 women, 32 men, 1 who did not indicate gender) enrolled in academic ESL courses. Multiple regression was used to investigate whether learner characteristics predicted students' progress in their ESL courses over the course of one semester. On average, employed students showed less improvement in their writing than students who did not work. Moreover, among employed students, those who cared for children under 18 and used more English at work tended to show smaller gains in their writing scores over the course of the semester, while those who used more English at home showed larger gains. Among unemployed students, more years of English study in the home country and planning to return home were associated with greater gains. In addition, those who planned to return to the home country were less likely to be anxious about using English and to believe that they were too old to learn English well. The article addresses possible reasons for, and practical implications of, these findings and suggests directions for further research.