The purpose of this study was to identify significant differences in academic achievement among virtual students of various backgrounds, demographics, and virtual learning environments. The study also sought to identify factors that may predict the academic achievement, as defined by final course grade, of virtual students. This study examined those relationships for the 476 students enrolled in virtual courses between January 2010 and January 2013 in Sullivan County Schools, TN. These students were in grades 7-12 during the time the courses were taken. Independent variables in Phase I of the study included gender, race/ethnicity, socio-economic status, prior number of virtual courses completed, and existing student grade point average. Independent variables in Phase II of the study included instructional dialogue in the virtual course, structure of the virtual course, and autonomy of the learner allowed in the virtual course. The researcher investigated the relationships between these independent variables and the dependent variable, academic achievement, as determined by final virtual course grade. The statistical methods used to answer the research questions included bivariate correlations, independent samples t-tests, and bivariate regression analysis. Two of the independent variables in Phase I of the study were found to be significant. Students identified as being economically disadvantaged tended to perform better academically in virtual courses than students identified as non-economically disadvantaged, as determined by final virtual course grade. A statistical significance was also found between existing student GPA and academic achievement in virtual environments. Students with a higher GPA prior to taking a virtual course tended to receive higher grades than those with lower existing GPAs. Using bivariate regression, existing GPA accounted for 25% of the variance in student academic achievement in virtual courses. All three of the independent variables in Phase II of the study were found to have a significant relationship with student academic achievement as determined by final virtual course grade. Students who reported high levels of instructional dialogue (frequency of teacher-student interactions, teaching presence, content interactions) tended to perform significantly higher than those reporting lower levels of instructional dialogue. Students who reported high levels of structure (instructional support, navigation, course design) tended to perform significantly higher than those reporting lower levels of structure in the course. Students who reported higher levels of autonomy (student ability to determine goals, learning experiences, and evaluation decisions) tended to perform significantly better academically than those who reported lower levels of autonomy.