The purpose of this study was to provide a structural model of the relationship between personality traits, intelligence beliefs and meta-cognitive awareness with test anxiety by mediating the educational self-efficacy in male students. This research was a correlation research and structural equation model. The statistical population consisted of all secondary school students in Bojnourd who were studying in the academic year of 1399- 1395. The multi-stage cluster sampling method was used. The test anxiety questionnaire (Abolghasemi et al., 1996) was designed to measure the variables The questionnaire of intelligence beliefs by Abdolfattah and Yeats (2006), the test anxiety questionnaire (TAI), NEO-FFI, metacognitive awareness scale (Mokhtari and Richard, 2002), Morgan and Jinns's academic self-efficacy scale (1999) used. The results showed that all factor loads of the markers are meaningful on their own variables. Also, all of the direct path coefficients of the exogenous variables including personality traits, intelligence beliefs and metacognitive awareness to academic self-efficacy, and academic self-efficacy to test anxiety were significant at the alpha level of 0.05. In addition, the model could account for more than 58% of academic self-efficacy and 30% of test anxiety changes. Also, the indirect effects of personality traits, metacognitive awareness and intelligence beliefs on test anxiety were significant through academic self-efficacy.