Block-design construction tasks reliably assess cognitive deficits due to brain injury. We examined the important aspects of this task with four experiments using normal subjects. Two problem-solving strategies are identified: an analytic strategy in which subjects mentally segment each block in the design to be constructed and a synthetic strategy, which involves wholistic pattern matching. Three experiments found a predominate analytic strategy. The time to place a single test block in a display decreased the greater the number of interior edges for that block in the design. Also, two-colored blocks requiring an orientation judgment were placed slower than solid blocks. The fourth experiment predicted overall construction times for a design from the number of solid blocks and interior edges of its blocks. These studies suggest refinements in the blockdesign test for investigating constructional disability in brain-damaged patients. We recommend such analyses of other neuropsychological tests.