In order to compare the capacities of a variety of compounds to interfere with P-glycoprotein (Pgp) function, a novel assay was set up to work on a large screening scale. The model assay measures the capacity of parental sensitive (Par) and multidrug-resistant (MDR) cells to efflux a small fixed amount of acetoxymethyl calcein (calcein-AM) after their pretreatment with concentration ranges of known Pgp modulators. This microplate cytometry-based assay was performed with two different pairs of cell lines, the human lymphocytic leukemia CEM cells and the murine monocytic leukemia P388 cells. For a given Pgp-expressing MDR cell line, a Pgp modulator EC50 was defined as the concentration required to restore half of the calcein retention shown by similarly treated Par cells. With both MDR-P388 and MDR-CEM cells, EC50 comparisons ranked five reference Pgp modulators as follows: SDZ 280-446 > SDZ PSC 833 > cyclosporin A > verapamil > vinblastine. Further use of the MDR-CEM cells could rank 15 Pgp modulators for their capacity to interfere with calcein-AM efflux as follows: SDZ 280-446 1.9 x > SDZ PSC 833 8.3 x > cyclosporin A 3.8 x > amiodarone 1.1 x > quinacrine 1.6 x > verapamil 1.4 x > quinidine 1.1 x > vinblastine 11 x > vincristine 2 x > chloroquine > beta-lumicolchicine > or = gamma-lumicolchicine > or = colchicine > etoposide > or = doxorubicin. This calcein-AM assay should open the way for ranking large numbers of novel structures for their potential Pgp modulator properties, particularly for an efficient screening of Pgp function antagonists, but it does not allow defining whether their inhibition may be competitive or not.