ABSTRACT In educational psychology and first language writing, many studies have established a link between self-regulation, key motivational factors associated with self-regulation, and academic achievement, but only a handful of research has examined this relationship in the context of L2 first-year writing. Using a sample of 170 non-native English speakers enrolled in different sections of two levels of first-year writing courses at an American university in Japan, the present study tested a hypothesized model of L2 first-year writing achievement using Structural Equation modeling. The model examined the impact of SRL strategy use, Academic Writing Self-Efficacy, Goal Orientation, Writing Attitude—key motivational factors associated with self-regulated learning—and general English proficiency, as measured by TOEFL iBT, on the writing achievement of students’ final essays, measured by their essay grades. The impact of gender on the hypothesized model of writing achievement was also investigated by conducting two separate Structural Equation modeling analyses on the hypothesized model for males and females. The study also examined the impact of SRL Strategy Use and Academic Writing Self-Efficacy on four levels of Writing Achievement, as well as the impact of four levels of Writing Experience on SRL Strategy Use and Academic Writing Self-Efficacy. The results of the study indicated that the hypothesized model had adequate fit to the data, and was, therefore, interpreted as being representative of the sample population examined in the current study. Statistically significant relationships in the model were found among the following variables: (a) English Proficiency and Essay Grade, (b) Academic Writing Self-Efficacy and Essay Grade, (c) Writing Attitude and Academic Writing Self-Efficacy, (d) Mastery Goal Orientation and SRL Strategy Use. These results corroborate findings in first-language and second-language writing research, which have reported statistically significant positive relationships among these variables, and lend support to the notion emphasized in socio-cognitive models of SRL that self-efficacy is a strong predictor of writing achievement. However, statistically significant relationships were not found among: (a) SRL Strategy Use and Essay Grade, (b) Writing Attitude and SRL Strategy Use, (c) Academic Writing Self-Efficacy and SRL Strategy Use, (d) Academic Writing Self-Efficacy and Goal Orientation. Possible explanations for the lack of statistically significant findings among the relationships between SRL and the other variables were attributed to the small sample size, and methods used to assess the use of SRL strategies. While the importance that the participants’ attribute to earning credits for the courses, as opposed to mastering writing skills, was considered a reason for the lack of a statistically significant relationship between Academic Writing Self-Efficacy and Goal Orientation. The examination of the influence of Gender on the hypothesized model of writing achievement indicated that the model for females had more adequate fit to the data than the model for males, suggesting that the model was more representative of the female participants. Differences in the models were found in the relationships between English Proficiency and SRL Strategy Use and the relationship between Academic Writing Self-Efficacy and Essay Grade. The results were in line with previous findings that have reported that female students use more SRL strategies and hold higher self-efficacy beliefs than male students. With regards to the influence of SRL Strategy Use and Academic Writing Self Efficacy on levels of Writing Achievement that ranged from Poor to Excellent, statistically significant differences were only found between the mean scores of the Poor and Excellent groups with regards to Writing Self-Efficacy. SRL Strategy Use did not exert a statistically significant difference on the mean scores of the groups. The results were in line with previous findings that reported the predictive influence of self-efficacy on writing achievement, but the results did not corroborate previous findings in relation to the predictive strength of SRL Strategy Use. The methodology used to assess the use of SRL strategies in the present study was considered a possible explanation for the lack of statistically significant results. In relation to the influence of Writing Experience on SRL Strategy Use and Academic Writing Self-Efficacy, the results also yielded non-significant differences between four groups with different levels of Writing Experience and SRL Strategy Use. This result was attributed to the broad nature of the method used to assess Writing Experience in the current study. Statistically significant differences were found between Academic Writing Self-Efficacy and Writing Experience, and the results supported previous findings in first language writing research, which have shown that learners with less experience often report higher levels of efficacy due to perhaps to overestimation of their skills. Overall, in the current study SRL did not predict the participants’ essay grades, and did not mediate the influence of other variables on essay grade. However, Academic Writing Self-Efficacy emerged as a powerful predictor of Essay Grades, and writing achievement. Therefore, while the current study supported social cognitive views about the predictive nature of self-efficacy on writing achievement, it did not corroborate theoretical assumptions about the relationship between the use of SRL strategies and writing achievement. / Teaching & Learning