This paper reports on a study that examines the effects of problem context on students' performance. The performance of 151 Year 10 students on six mathematical problems was compared with the performance on fifteen variants with more and less context familiarity (CF) and engagement (CE) across levels of context use (LCU). The latter explanatory variables (CF, CE, and LCU) are used to estimate the strength of the relationship among them and the students' performance. Results show that neither CF nor CE affect students' performance but LCU demanded in solving a problem does.