As an affective factor, test-taking anxiety has been investigated in different contexts in the<br />past two decades. However, the mixed results of the relationship between test-taking anxiety and<br />L2 learners&rsquo; test performance show that the instrumentation for the assessment of test-taking<br />anxiety and the factors comprising the construct of test-taking anxiety trait requires more<br />investigation in order to shed more light on the issue. To this end, a test-taking anxiety<br />questionnaire (Sarason, 1975) [27] and a general English test were administered to 164 ESP<br />students of Engineering enrolled in a B.A. program to document (a) the degree of their test taking<br />anxiety, (b) the relationship between test-taking anxiety and test performance, and (c) the factor<br />loadings of anxiety based on exploratory factor analysis. The results show that L2 learners&rsquo; test<br />anxiety is rather low, with most of its components having no significant negative correlation with<br />test performance. The results of exploratory factor analysis reveal the loading of test anxiety trait<br />on the rather overlapping three factors of specific test anxiety, general test anxiety, and test preparation anxiety. However, out of these factors, general test anxiety, due to its functioning at the higher-order affective level, has a significant negative correlation with test performance. By contrast, test preparation anxiety, in view of facilitating test performance, manifests a positive, albeit non-significant, correlation with test performance. The results have two implications: (a) as the correlations and loadings on test anxiety factors proved to be of both negative and positive types, the anxiety questionnaire is not monolithic and hence it is not a proper measure in case the linear relationship between test anxiety and test performance is the focus of the study; and (b) test anxiety does not seem to much influence on test performance at the micro- test-specific level.