In England, students obtain General Certificate of Secondary Education (GCSE) qualifications, typically at age 16. Certain GCSEs are tiered; students take either higher-level (higher tier) or lower-level (foundation tier) exams, which may have different educational, career and psychological consequences. In particular, foundation tier entry, if inappropriate, risks capping students' achievement because of the restricted range of attainable grades and reduced learning that may occur. Tiering decisions may be affected by other aspects of the education system in which they take place, such as by the timing of assessment. The move to linear assessment in 2012 provided a unique opportunity to compare tiering decisions for the same GCSE specifications when taken in a linear system, where students are exclusively assessed at the end of the course, with tiering decisions in a modular system, where students are assessed at different time points. Multilevel logistic regression was used to examine students' likelihood of being entered for the foundation tiers of GCSEs in science, language and mathematics in two exam sessions: June 2013, which allowed modular assessment, and June 2014, which required linear assessment. The analyses also investigated whether these effects depended on student characteristics. Results showed that foundation tier entry was less likely in the linear than modular system for GCSEs in science and languages, but more likely for one mathematics GCSE. This pattern contrasts with concerns that linear assessment may encourage general risk-aversion, and instead indicates that effects on tiering decisions are more complicated, varying by subject and student factors.