This paper reports the effects of game technology on student learning in mathematics as investigated in two data sets collected from slightly different subjects. In the first, 41 second graders (7 or 8 years old) from two classes used either a technology-based game or a paper-based game for 5 weeks. For the next 13 weeks, both classes used a technology-based game either two times per week, or more than three times per week. A quasi-experimental control-group design with repeated measures analysis of variance and analysis of covariance was employed to explore performance differences between groups. The second study examined student learning in relation to characteristics such as their game performance, attitudes toward the game and toward mathematics, and gender and ethnicity. During a 4-month period, 50 second grade students from three classes played a technology-based game under conditions that varied depending on their teacher's direction. Multiple regression was used to determine the relationship between students' arithmetic scores and learner characteristics. Descriptive analyses by ability level, gender and ethnicity, and interview data about attitudes toward the technology game were also analyzed. Results from the two studies revealed that using a technology-based game in the classroom was beneficial to students of all ability levels in learning arithmetic skills. (Contains 11 tables and 2 figures.)