A computerized block design task was developed which records temporal and nontemporal measures of performance. This study evaluates the reliability of the measures and reports their intercorrelations. With one exception, the measures showed moderate to good reliability. The results indicate that increasing the difficulty of the task and testing a more diverse sample may be necessary for improved reliability. A nontemporal method of scoring a block-design task would be useful when testing persons who have handicaps affecting motor skills, but no central nervous system deficits.