Jensen (1969) proposed a two-factor model of intelligence by which mental abilities can be divided into memory and higher mental processes labeled "Level I" and "Level II" respectively. This study employed statistical methods characteristic of construct validation to examine a battery of memory, intelligence, and achievement tests, seeking evidence of the existence, convergent validity and discriminant validity of the constructs of Level I and Level II intelligence. A principal components factor analysis followed by a promax rotation to oblique simple structure was performed, following Jensen's methodology. To examine the convergent and discriminant validity of each of the constructs, Campbell and Fiske's multitrait-multimethod matrices were constructed. The methods involved were group vs. individual administration for the first matrix and verbal vs. nonverbal response mode for the second. Each of these analyses was performed first using the entire battery of 15 variables. The nine variables used in the second set of analyses were those judged to be "purer" measures of Level I and Level II intelligence. Rural Caucasian fifth grade children were tested. The first factor analysis using all 15 variables yielded six factors providing little support for the two-factor theory. Using all 15 variables there was little support for either the convergent or discriminant validity of the constructs. The factor analysis using the "purer" measures yielded three factors, again providing little support for the theory. Examination of the multitrait-multimethod matrices provided little support for the convergent or discriminant validity of the constructs. In conclusion, the study calls into question the existence of the constructs of Level I and Level II intelligence. In particular, there was little support for the existence of Level I intelligence as a coherent psychological trait. (Author/RC)