This study was designed to formulate a predictive equation to identify community college biology students (N=420) who most probably would not succeed in science courses. A College Biology Student Survey (developed for the study), Nelson Denny Reading Test (Form-F), College Guidance Placement (CGP) Arithmetic Test, and An Inventory of Piaget's Developmental Tasks were used to obtain student background characteristics, reading/mathematics ability, and cognitive developmental level. Data collected from students in 22 biology classes indicated: more females (82%) than males (22%); a mean age of 26; racially balanced classes (49% white and 51% non-white); that 71% were single or formerly married; that 9% were at the formal level of cognitive development; a mean score of 15/25 on the mathematics test (measuring 6th-grade mathematics ability); an average 12th-grade reading level; and an average course grade of 2.7(C+). When data were analyzed using multiple regression, background characteristics and reading/mathematics ability accounted for 42% variance in cognitive development and 43% variance in achievement; cognitive development accounted for 12.8% variance in achievement; and with cognitive development as a mediating variable, background characteristics and reading/mathematics ability accounted for 22% variance in achievement. Recommendations were made on the predictive equation formulated from these results. (Author/JN)