The passive transport of calcium and cobalt and their interference were studied in human red cells using (45)Ca and (57)Co as tracers. In ATP-depleted cells, with the ATP concentration reduced to about 1µM, the progress curve for (45)Ca uptake at 1mM rapidly levels off with time, consistent with a residual Ca-pump activity building up at increasing [Ca(T)](c) to reach at [Ca(T)](c) about 5µmol(lcells)(-1) a maximal pump rate that nearly countermands the passive Ca influx, resulting in a linear net uptake at a low level. In ATP-depleted cells treated with vanadate, supposed to cause Ca-pump arrest, a residual pump activity is still present at high [Ca(T)](c). Moreover, vanadate markedly increases the passive Ca(2+) influx. The residual Ca-pump activity in ATP-depleted cells is fuelled by breakdown of the large 2,3-DPG pool, rate-limited by the sustainable ATP-turnover at about 40-50µmol(lcells)(-1)h(-1). The apparent Ca(2+) affinity of the Ca-pump appears to be markedly reduced compared to fed cells. The 2,3-DPG breakdown can be prevented by inhibition of the 2,3-DPG phosphatase by tetrathionate, and under these conditions the (45)Ca uptake is markedly increased and linear with time, with the unidirectional Ca influx at 1mM Ca(2+) estimated at 50-60µmol(lcells)(-1)h(-1). The Ca influx increases with the extracellular Ca(2+) concentration with a saturating component, with K(½(Ca)) about 0.3mM, plus a non-saturating component. From (45)Ca-loaded, ATP-depleted cells the residual Ca-pump can also be detected as a vanadate- and tetrathionate-sensitive efflux. The (45)Ca efflux is markedly accelerated by external Ca(2+), both in control cells and in the presence of vanadate or tetrathionate, suggesting efflux by carrier-mediated Ca/Ca exchange. The (57)Co uptake is similar in fed cells and in ATP-depleted cells (exposed to iodoacetamide), consistent with the notion that Co(2+) is not transported by the Ca-pump. The transporter is thus neither SH-group nor ATP or phosphorylation dependent. The (57)Co uptake shows several similarities with the (45)Ca uptake in ATP-depleted cells supplemented with tetrathionate. The uptake is linear with time, and increases with the cobalt concentration with a saturating component, with J(max) about 16µmol(lcells)(-1)h(-1) and K(½(Co)) about 0.1mM, plus a non-saturating component. The (57)Co and (45)Ca uptake shows mutual inhibition, and at least the stochastic Ca(2+) influx is inhibited by Co(2+). The (57)Co and (45)Ca uptake are both insensitive to the 1,4-dihydropyridine Ca-channel blocker nifedipine, even at 100µM. The (57)Co uptake is increased at high negative membrane potentials, indicating that the uptake is at least partially electrogenic. The (57)Co influx amounts to about half the (45)Ca influx in ATP-depleted cells. It is speculated that the basal Ca(2+) and Co(2+) uptake could be mediated by a common transporter, probably with a channel-like and a carrier-mediated component, and that (57)Co could be useful as a tracer for at least the channel-like Ca(2+) entry pathway in red cells, since it is not itself transported by the Ca-pump and, moreover, is effectively buffered in the cytosol by binding to hemoglobin, without interfering with Ca(2+) buffering. The molecular identity of the putative common transporter(s) remains to be defined.