Cloze was officially introduced in a journal on Journalism as a technique for estimating text readability and as "a new psychological tool for measuring the effectiveness of communication" (Taylor, 1953: 415). Different varieties of cloze have since been developed and experimented upon as measures of such diverse traits as reading comprehension and language proficiency. The findings of numerous corrleational studies on cloze as a measure of either skill is at best unsatisfactory and indeed contradictory. The present study seeks to find an answer to the question of whether standard cloze (with different text difficulty levels) is a valid measure of EFL reading comprehension (with IELTS Reading Paper as the criterion). 76 junior and senior students majoring in English Language and Literature at Urmia University participated in the study, where they sat 3 versions of standard 5-th deletion rate cloze tests as well as the Reading Paper of an Institutional IELTS (UCLES, 1995, 1997). While the results are in accordance with most previous research findings that cloze is a valid measure of EFL reading comprehension, serious problems are identified and discussed on the appropriacy of such a validation technique as correlation.