Interspersing easy problems among challenging problems has been shown to both increase students' preference for completing math worksheets and the fluency with which they complete challenging problems. Although early research examining the interspersal procedure in mathematics was conducted with college-age students, there are a growing number of studies utilizing the interspersal procedure in elementary and secondary school settings. A particular concern with existing research however is whether the beneficial effects are maintained across multiple sessions and whether effects generalize when implemented with large groups of students. The current study addressed this concern by administering both control and interspersal math worksheets consisting of instructionally relevant problems to 75 fourth grade students across four sessions. Because students' instructional levels differed, analyses were conducted separately for students identified as fluent and less fluent based upon student performance on the control worksheet administered during session 1. Dependent measures included digits correct and problems correct as well as students' responses to four preference questions following each session. Although results differ from previous research in that students completed significantly more digits correct on control worksheets, explanations of these differences are examined and implications are discussed.