High school students who spend long hours in paid employment during the school year are at increased risk of lower grades and higher substance use, although questions remain about whether these linkages reflect causation or prior differences (selection effects). Questions also remain about whether such associations vary by socioeconomic status (SES) and race/ethnicity. This study examines those questions using nationally representative data from two decades (1991-2010) of annual Monitoring the Future surveys involving about 600,000 students in 10th and 12th grades. White students are consistently more likely than minority students to hold paid employment during the school year. Among White and Asian American students, paid work intensity is negatively related to parental education and grade point averages (GPA) and is positively related to substance use. Also among Whites and Asian Americans, students with the most highly educated parents show the strongest negative relations between work intensity and GPA, whereas the links are weaker for those with less educated parents (i.e., lower SES levels). All of these relations are less evident for Hispanic students and still less evident for African American students. It thus appears that any costs possibly attributable to long hours of student work are most severe for those who are most advantaged--White or Asian American students with highly educated parents. Working long hours is linked with fewer disadvantages among Hispanic students and especially among African American students. Youth employment dropped in 2008-2010, but the relations described above have shown little change over two decades.