A group of adult education teachers sought to determine if a readability formula applied to classroom materials provided a match with student reading levels. Cloze tests were used to determine whether students were able to understand the material. In three of the eight samples, the Cloze test showed that the readability test did not accurately indicate material suitability. (Teachers reported research separately because of difficulty in deciding how to present their findings.) The teachers concluded that constructing Cloze tests was a better use of time than figuring out readability using formulas. They also found some evidence that students could read better when using materials in which they were interested. They recommended that, if mixed reading levels occur among students in adult education classes, teachers should use materials with reading levels corresponding to those of the lower readers. Students with higher reading levels could still learn from these materials, but too-high levels would frustrate the slower students and possibly cause them to drop out of adult education programs. (Author/KC)