This study investigated the influence of student and classroom characteristics on a sample of 17-year-old students' (N=1955) inquiry ability. The sample was obtained from a 1981/1982 national assessment in science carried out by the Minnesota Science Assessment and Research Project. Specific areas addressed included: (1) the effectiveness of the Model of Educational Productivity (MEP) for predicting students' inquiry skills; (2) whether predictors of science inquiry skills differed for males and females; and (3) whether predictors of science inquiry skills differed for white and nonwhite students. Results indicated that the model was capable of accounting for between 24 and 32 percent of the variance in inquiry skill for the general population of 17-year-olds; more specifically, ability alone accounted for between 17 and 22 percent of the variance in inquiry skill for the general population. Little difference in the prediction of inquiry skill for males and females using the model was found. In addition, for nonwhite students, the MEP accounted for only 18 percent of the variance in science inquiry skill; specifically, ability accounted for only 6 percent of the variance, suggesting that little is known about the factors contributing to the science inquiry skill of nonwhite students. (JN)