Various modifications of the Boyden chamber chemotaxis assay have been used to screen patients for abnormalities in granulocyte or monocyte motility. In most cases, cell motility has been assessed by quantitating the fraction of cells that migrates from an upper chamber through a filter toward a lower chamber containing chemoattractant. Existing versions of the assay have several shortcomings. They are labor-intensive, require relatively large numbers of cells and lengthy incubation, or they require visual cell counting and do not permit assessment of cells which may drop off the filter into the attractant medium. We have improved the accuracy and efficiency of existing microchamber assays by using 51Cr-labeled cells to eliminate microscopic cell counting, shortening the incubation time, adjusting the assay sensitivity, and accounting for cells which drop off into the attractant well. The modified method uses Neuroprobe multiwell microchambers and two 10 microns polycarbonate filters with 3 microns pores on top of one 100 microns nitrocellulose filter. The optimal incubation period is 60 min, and the assay requires about one-fifth as many cells as the standard Boyden chamber methods. Cell drop-off can be measured accurately by harvesting the attractant wells with detergent, and the assay sensitivity is comparable to that of existing radiometric assays using large chambers. The data indicate that the range of chemotactic and random motility of normal granulocytes and monocytes measured in the modified assay system is comparable to that reported for studies which have used established motility assays.