Two studies investigated the relationships between readability, prior knowledge, and interest in the material being read. In the first study, 169 college freshmen read cloze passages of materials from the Nelson-Denny Reading Tests, then one week later read the whole passages and completed the related multiple choice questions. A control group merely read the passages and answered the multiple choice questions. The cloze scores correlated with test passage readability, but the multiple choice scores did not correlate with the readability scores. When data from 50 more subjects who took the multiple choice tests without reading the passages were entered into the calculations much higher correlations were found between these "corrected" scores and the readability scores. In the second study, 266 entering freshmen supplied reading interest data and read high and low interest passages with manipulated reading difficulty levels. The results showed that prior knowledge and reading interest were significantly related. Further analyses of data from 18 subjects selected for combinations of interest and readability revealed that prior knowledge was the more powerful variable in the study. The collective results from the two studies indicated that readability measures tend to overestimate passage difficulty when prior knowledge and reading interest are high, but underestimate passage difficulty when these variables are low. (RL)