This document reviews studies that investigated certain inadequacies of English as a Second Language (ESL) reading comprehension tests. The first study examined whether three ESL reading comprehension tests required information principally from the text or from background knowledge. Results indicated that background knowledge differentially affects reading comprehension at different levels of proficiency and that a large proportion of the variance in reading comprehension scores can be accounted for by the reader's background knowledge. The second study measured reasoning in the Test of English as a Foreign Language (TOEFL) reading comprehension subtest. Results indicated that the subtest seemingly lacks validity. The third study investigated whether an ESL test distinguished between the subskills it professed to measure. Results indicated that only 44% of the total test exhibited internal construct validity. The fourth study examined behavioral anchoring, whose purpose is to define achievement and to make test scores interpretable in terms of what students at different ability levels could and could not do. Results indicated that higher ability level students excelled with items which depended on implicit information; that all proficiency levels exhibited achievement with linguistic structures which related parts of the text to one another; and that the derivational complexity and the readability level estimates co-varied with an increase in proficiency level. Specific remedies for each of the weaknesses of ESL reading comprehension tests are suggested. (Twenty-six references are appended.) (RS)