Over the last decade, the idea of providing students with choices in their learning experience has attracted academic interest (Flowerday & Shraw, 2000; Katz & Assor, 2007; Patall, Cooper, & Robinson, 2008). Although some previous research has suggested that choice is beneficial to language learning, other research has indicated that choice has negligible (Iyengar & DeVoe, 2003) or even damaging effects (D’Ailly, 2004; Stefanou, Perencevich, DiCintio, & Turner, 2004) on language acquisition. Considerable differences in the focuses of previous research can explain the conflicting results of these choice studies (Iyengar & DeVoe, 2003; Schwartz, 2004); however, researchers agree that choice is closely associated with motivation (Stefanou et al., 2004). For instance, various motivational models, such as self-determination theory (Deci & Ryan, 2000), include the concepts of choice, autonomy, and control as key elements of intrinsic motivation and performance. This study had three main purposes, the first of which was to quantitatively examine the relationships among Reading Anxiety, Autonomy, Interest, Reading Self- Efficacy, and Reading Proficiency in Japanese EFL students in a first-year pre-intermediate reading course. The second purpose was to quantitatively examine the effect of having No Choice, Option Choice, and Active Choice (Reeve, Nix, & Hamm, 2003) on Reading Anxiety, Reading Autonomy, Reading Interest, Reading Self-Efficacy, and reading performance over one academic year in a foreign language reading curriculum. The final purpose was to qualitatively corroborate and support the quantitative findings through a series of structured interviews based on students’ beliefs and attitudes toward the provision of choice in the reading curriculum. A quantitative quasi-experimental design supported by a qualitative phenomenological component was used during the year-long longitudinal study with 201 first-year Japanese EFL students at a private university in Japan. Nine intact classes were randomly assigned into three groups: No Choice (n = 66), Option Choice (n = 67), and Active Choice (n = 68), as defined by Reeve et al. (2003). Affective Variable Questionnaires were administered to measure the levels of Reading Anxiety, Reading Autonomy, Reading Interest, and Reading Self-Efficacy before, during, and after a 32-week treatment. The results of reading performance measures, including Vocabulary Definition and Vocabulary in Context quizzes, Intensive Reading tests, Extensive Reading quizzes, Timed Reading assignments, Composite TOEFL, and TOEFL Reading component scores were tracked over the academic year. The results showed low to medium Pearson correlations ranging between r = - .39 to r = .29 among Reading Anxiety, Reading Autonomy, Reading Interest, and Reading Self-Efficacy. In addition, a stable, significant relationship was found between Reading Self-Efficacy and Reading Proficiency, as measured by students’ TOEFL scores and TOEFL Reading Component scores at the start and end of the academic semester. Initially, no such relationship was found between Reading Anxiety, Reading Autonomy, Reading Interest, and Reading Proficiency. However, by the end of the academic year, significant correlations were found among the Reading Autonomy, Reading Interest, Composite TOEFL, and TOEFL Reading component scores. The results indicated significant changes in the affective variables within each group over the academic year. Over the year, significant decreases in Reading Anxiety, and significant increases in Reading Self-Efficacy in each of the three groups were particularly salient. In addition, there were significant changes in many of the Reading Performance measures for each of the groups; however, only the Active Choice group had significant changes in all seven Reading Performance measures over the year. In terms of the effect of choice on the affective variables, students in both the Active Choice and the Option Choice groups had significantly higher Autonomy gains than students in the No Choice group over the academic year. Thus, giving students any type of choice in their reading curriculum exerted a positive effect on Reading Autonomy. With regards to the effect of choice on reading performance, mixed results were found in the reading components among the three groups. First, in the Intensive Reading and Timed Reading components, students in the Active Choice group performed significantly better than students in the Option Choice and No Choice groups. This finding indicated that when choice is given to students, it is necessary that the locus of control be with the student. With respect to Vocabulary Definitions and Vocabulary in Context components, both the Active Choice and Option Choice groups had significantly higher scores than the No Choice group. In other words, any choice was considered better than no choice. The type of choice had no effect on the Vocabulary components. In Extensive Reading, the Active Choice group significantly outperformed the No Choice group in the Extensive Reading quizzes; however, the Option Choice group was not significantly different from the other two groups. The results indicated that only autonomous choice led to greater self-determination, and increases in performance. Finally, no differences were found among the three groups in the Composite TOEFL scores and the TOEFL Reading component scores. The quantitative findings were corroborated by interviews with 18 students with a wide range of motivation and reading performance, as measured by the Affective Variables Questionnaire and the reading performance measures. The students were interviewed about the treatment process and their feelings about having choice in the reading curriculum. Common themes derived from the interview data indicated that choice affected students’ sense of Reading Autonomy. A common pattern emerged from the data indicating that students in the Active Choice group with lower levels of affect and reading performance were less comfortable making choices than students with higher levels of affect and reading performance abilities. Additionally, students in the No Choice group with higher levels of affect and reading performance were frustrated by the lack of choice in the reading course. The study contributed four unique points to the field of choice in language learning. First, choice was found to increase students’ sense of Reading Autonomy, a key component in intrinsic motivation and successful learning (Littlewood, 1999). Next, having any type of choice was found to be beneficial in Vocabulary acquisition. Moreover, only autonomous choice was found to be advantageous in the more complex tasks of Intensive Reading, Extensive Reading, and Timed Reading. Finally, the benefits of choice did not extend to performance on the Composite TOEFL and TOEFL Reading components. The testing environment and the lack of choice available in standardized testing were demotivating and contributed to a decrease in reading performance. The mixed results of this study indicate that choice is a complex phenomenon. The field of choice in education and language learning offers a wealth of teaching and research possibilities for future study. / Teaching & Learning