Large numbers of differential counts are generated in hospital laboratories by physician requesting repeat counts, often daily. Their purpose is to detect trends in polymorphonuclear neutrophils (PMN) to monitor infections or therapy with marrow-depressant drugs. Recently, a number of automated differential counters have come on the market to handle the still increasing load of differentials. We proposed that a strategy of performing differential counts only after a significant change in total WBCs had taken place would result in little or no loss of information. Since WBC counts are routinely performed by automated instruments already available in most laboratories, the purchase of expensive differential counters might thus be avoided. To test this thesis, we analyzed 415 successive sequences of WBC and differential counts, each sequence consisting of four determinations within a 10-day period. Criteria for significant trends in WBC and PMN were based on clinical judgment or on recently determined day-to-day physiologic variations. Among 415 sequences examined, only four could be found in which PMN had changed without a concomitant change of equal significance in WBC. It was concluded that the proposed strategy of using WBC as an indicator of developing neutropenia or neutrophilia was sound. It would have resulted in loss of information in about 1% of instances only, at a considerable saving in differential counts.