It is well established that socioeconomic status, cognitive ability, and non-cognitive abilities such as self-efficacy are substantially associated with academic achievement. However, the specific relationships of these variables remain a relatively unexplored topic in regard to more recent and representative samples. The current study examined such relations by taking advantage of two cohorts (total N = 12,315) of Swedish students at the elementary (Grade 6) and lower-secondary school levels (Grade 8) in the compulsory school. The regression models showed that all three variables explained a substantial portion of grade variance, with cognitive ability having the strongest relationship, followed by non-cognitive abilities, and SES. Longitudinal associations, which accounted for previous academic achievement, showed that the three variables did still explain a substantial amount of grade variance.