Standard models for carrier-mediated nonelectrolyte transport across cell membranes do not explain sugar uptake by human red blood cells. This means that either (1) the models for sugar transport are incorrect or (2) measurements of sugar transport are flawed. Most measurements of red cell sugar transport have been made over intervals of 10 s or greater, a range which may be too long to measure transport accurately. In the present study, we examine the time course of sugar uptake over intervals as short as 5 ms to periods as long as 8 h. Using conditions where transport by a uniform population of cells is expected to be monophasic (use of subsaturating concentrations of a nonmetabolizable but transported sugar, 3-O-methylglucose), our studies demonstrate that red cell sugar uptake is comprised of three sequential, protein-mediated events (rapid, fast, and slow). The rapid phase is more strongly temperature-dependent than the fast and slow phases. All three phases are inhibited by extracellular (maltose or phloretin) or intracellular (cytochalasin B) sugar-transport inhibitors. The rate constant for the rapid phase of uptake is independent of the 3-O-methylglucose concentration. The magnitude (moles of sugar associated with cells) of the rapid phase increases in a saturable manner with [3-O-methylglucose] and is similar to (1) the amount of sugar that is retained by red cell membrane proteins upon addition of cytochalasin B and phloretin and (2) the d-glucose inhibitable cytochalasin B binding capacity of red cell membranes. These results are consistent with the hypothesis that previous studies have both under- and overestimated the rate of erythrocyte sugar transport. These data support a transport mechanism in which newly bound sugars are transiently sequestered within the translocation pathway where they become inaccessible to extra- and intracellular water.