In sickle cell anemia (SS), some red blood cells dehydrate, forming a hyperdense (HD) cell fraction (>1.114 g/mL; mean corpuscular hemoglobin concentration [MCHC], >46 g/dL) that contains many irreversibly sickled cells (ISCs), whereas other SS red blood cells dehydrate to an intermediate density (ID; 1.090 to 1.114 g/mL; MCHC, 36 to 46 g/dL). This study asks if the potassium-chloride cotransporter (K:Cl) and the calcium-dependent potassium channel [K(Ca2+)] are participants in the formation of one or both types of dense SS red blood cells. We induced sickling by exposing normal density (ND; 1.080 to 1.090 g/mL; MCHC, 32 to 36 g/dL) SS discocytes to repetitive oxygenation-deoxygenation (O-D) cycles in vitro. At physiologic Na+, K+, and Cl-, and 0.5 to 2 mmol/L Ca2+, the appearance of dense cells was time- and pH-dependent. O-D cycling at pH 7.4 in 5% CO2-equilibrated buffer generated only ID cells, whereas O-D cycling at pH 6.8 in 5% CO2-equilibrated buffer generated both ID and HD cells, the latter taking more than 8 hours to form. At 22 hours, 35% +/- 17% of the parent ND cells were recovered in the ID fraction and 18% +/- 11% in the HD fraction. Continuous deoxygenation (N2/5% CO2) at pH 6.8 generated both ID and HD cells, but many of these cells had multiple projections, clearly different from the morphology of endogenous dense cells and ISCs. Continuous oxygenation (air/5% CO2) at pH 6.8 resulted in less than 10% dense cell (ID + HD) formation. ATP depletion substantially increased HD cell formation and moderately decreased ID cell formation. HD cells formed after 22 hours of O-D cycling at pH 6.8 contained fewer F cells than did ID cells, suggesting that HD cell formation is particularly dependent on HbS polymerization. EGTA chelation of buffer Ca2+ inhibited HD but not ID cell formation, and increasing buffer Ca2+ from 0.5 to 2 mmol/L promoted HD but not ID cell formation in some SS patients. Substitution of nitrate for Cl- inhibited ID cell formation, as did inhibitors of the K:Cl cotransporter, okadaic acid, and [(dihydroindenyl) oxy]alkanoic acid (DIOA). Conversely, inhibitors of K(Ca2+), charybdotoxin and clotrimazole, inhibited HD cell formation. The combined use of K(Ca2+) and K:Cl inhibitors nearly eliminated dense cell (ID + HD cell) formation. In summary, dense cells formed by O-D cycling for 22 hours at pH 7.4 cycling are predominately the ID type, whereas dense cells formed by O-D cycling for 22 hours at pH 6.8 are both the ID and HD type, with the latter low in HbF, suggesting that HD cell formation has a greater dependency on HbS polymerization. A combination of K:Cl cotransport and the K(Ca2+) activities account for the majority of dense cells formed, and these pathways can be driven independently. We propose a model in which reversible sickling-induced K+ loss by K:Cl primarily generates ID cells and K+ loss by the K(Ca2+) channel primarily generates HD cells. These results imply that both pathways must be inhibited to completely prevent dense SS cell formation and have potential therapeutic implications.