Studies of change in neuropsychological function over time in both healthy and diseased populations have been hindered by the absence of alternate forms of most commonly used neuropsychological tests. The Penn Conditional Exclusion Test (PCET) was developed as a new cognitive instrument to assess "executive" functioning with four alternate forms. In Study 1, the PCET was administered in counterbalanced order to 80 healthy young adults to establish equivalent test difficulty. Results revealed that the four versions were related on categories achieved, total number of errors and total number of trials. In Study 2, the PCET was administered to 25 healthy adults of a wide age range with a battery of computerized neuropsychological tests to assess construct validity. Convergent validity was confirmed by positive correlations between the PCET and a measure of abstraction. Divergent validity was established through low, nonsignificant correlations between the PCET and measures of facial emotion recognition, word and face memory, visuospatial function, and verbal reasoning.