Earlier studies from our laboratory (Dembo, M., Sirotnak F. M., and Moccio, D. M. (1984) J. Membr. Biol. 78, 9-17) suggested that methotrexate (MTX) efflux from L1210 cells was mediated predominantly by an ATP-dependent, outwardly directed, mechanism. To examine this process further, we utilized predominantly (74%) inside-out plasma membrane vesicle preparations derived from an L1210 cell variant (L1210/R24) with 15-fold reduced Vmax for [3H]MTX influx. Efflux of [3H]MTX, under nonionic buffer conditions, in these inside-out membrane vesicles was temperature and ATP dependent (apparent Km = 0.40 +/- 0.06 mM), osmotically sensitive, and unaffected by protonophores. The presence of K+, Na+, Cl-, and HCO3- at their physiological concentrations had no effect on [3H]MTX efflux. Other triphosphonucleotides (GTP and CTP), but not a nonhydrolyzable analogue, adenosine-5'-O-(3-thiotriphosphate) (ATP gamma S), could also stimulate efflux, but to a lesser extent. Also, ATP gamma S and orthovanadate were potent inhibitors of ATP-dependent efflux of [3H]MTX. Other experiments revealed a system with low saturability for [3H]MTX during efflux (apparent Km = 46 +/- 7 microM), but extremely high capacity (106 +/- 15 pmol/min/mg protein), and a pH optimum in the range of 5.5-6. However, appreciable efflux was measured in the physiological range of pH 6.7-6.9. A number of inhibitors or copermeants for ATP-dependent [3H]MTX efflux in intact L1210 cells were inhibitors of ATP-dependent efflux in inside-out plasma membrane vesicles, including, cholate, bromosulfophthalein, verapamil, quinidine, and reserpine. These findings and other results showing that bromosulfophthalein will completely inhibit efflux are consistent with a role for an ATPase in [3H]MTX efflux, and suggest that the process under study is the bromosulfophthalein-sensitive, ATP-dependent route responsible for the majority of [3H]MTX efflux in intact L1210 cells.