In this paper, we identify the contribution of differences in test effort to gender gaps and socioeconomic gaps in achievement. We leverage question response time and random question order to obtain causal estimates of the effect of student effort on performance. Subsequently, we evaluate how differences in performance change when students would have made equal time investments. We find that effort explains around 25 percent of the socioeconomic gap in math and reading. For gender, correcting for effort closes around 18 percent of the reading gap while it increases the advantage of boys in math. Looking at average achievement, gender differences in effort can explain 49 percent of the gender achievement gap. We also show that the returns to response time are strongly underestimated by fixed effects models.