Recent reports based on PISA data have shown a generally positive relationship between the amount of time spent on homework and achievement, and a negative relationship between a measure of socioeconomic status and homework time for secondary students. These findings suggest that homework practices are either reflecting or contributing to achievement differences based on a students' socioeconomic and cultural background. However, a research gap has been identified in relation to how other demographic and educational variables associated with differences in achievement relate to differences in homework practice. We examined relationships between student socioeconomic status, gender, language background, school type and school location with various measures of homework time and frequency of completion for mathematics homework; two measures of homework time and effort available in the 2012 PISA data. Analysis of Australian 2012 PISA data largely confirmed that factors associated with variation in achievement were generally also associated in measures of homework practice. Suggestions are made for further research to illuminate the processes by which the independent variables examined affect homework practice.