1. Acute experiments were carried out on anaesthetized dogs during metabolic alkalosis produced by I.V. administration of NaHCO(3). Partial constriction of one ureter led to a significant rise in the HCO(3) (-) threshold, beyond the simultaneous value for the other kidney. The magnitude of the increase was not correlated with the reduction of glomerular filtration.2. Stop-flow analysis, following complete unilateral obstruction of urine flow, demonstrated proximal as well as distal tubular reabsorption of bicarbonate. At any given plasma P(co2) the detailed configuration of the concentration changes which developed depended on (a) the presence and concentration of mannitol, (b) the duration of urinary stasis, and (c) the plasma concentration of HCO(3) (-).3. If a solution containing 15% (w/v) mannitol was infused I.V., the HCO(3) (-) concentration in free flow urine was lower than in plasma, and it fell further during arrest of flow in the entire column of trapped fluid. If less mannitol was infused, or none at all, interruption of urine flow led to a striking increase of HCO(3) (-) concentration in the distal portion of the occluded column, and to a fall in the fluid arrested in the proximal segments.4. It was demonstrated that the HCO(3) (-) concentration attained after 2(1/2), 6, or 15 min of urinary stasis at any point in the trapped fluid column was due to the combined effects of water reabsorption and HCO(3) (-) reabsorption which proceeded independently, and with a different time course.5. If mannitol was administered the lowest urinary HCO(3) (-) concentration in the series moved progressively to a more distal location with increasing duration of urinary stasis. When HCO(3) (-) concentration peaks were present in distal fluid they were conspicuous only after short interruptions of urine flow; during extended stop-flow periods they became attenuated, or disappeared. If no mannitol was administered this did not occur.6. Provided the plasma level of HCO(3) (-) was sufficiently elevated, mannitol (15%, w/v) was administered, and the time available for reabsorption was lengthened by ureter obstruction, much larger concentration differences between plasma and trapped fluid developed than the largest that are ever found between the plasma and freely draining urine. The magnitude of the largest plasma-urine (P-U) concentration difference for HCO(3) (-) increased with intratubular ;contact time', and no limiting value was found.7. Potassium concentration in distal occluded fluid fell with prolonged duration of stasis. This was related to the slow and progressive diminution of distal HCO(3) (-) concentration. But if instead of bicarbonate a nonreabsorbable anion, such as phosphate, was the dominant distal anion, K(+) concentration in distal fractions remained high and rose further with time.