The matching cloze format used in constructing the reading comprehension parts of the Micronesian Achievement Test Series (MATS) is described as a holistic approach which measures skills comparable to those needed by English as a second language students in actual reading situations. Kenneth Goodman's theory about the strategies used by good readers is described as it relates to the construction of matching cloze tests. Matching cloze requires the examinee to re-read and comprehend more of the passage than do multiple-choice cloze tests. Nine studies of the matching cloze procedure, conducted since its introducation in 1975, are summarized. They indicate that matching cloze is a reliable and valid method of testing reading achievement. The MATS reliability is noted to be quite high, from .82 to .96. The matching cloze procedure is also briefly compared with the standard cloze, multiple choice cloze, and traditional multiple choice tests; each method has advantages and disadvantages for different situations and with different samples of students. Tables are presented to summarize the nine published studies; reliability and validity estimates; and characteristics of different test formats. Instructions for test construction, administration, and scoring, as well as sample items, are appended. (GDC)