When normal people ingest 90 mEq/day of K+ in their diet, they absorb about 90% of intake (81 mEq) and excrete an equivalent amount of K+ in the urine. Normal fecal K+ excretion averages about 9 mEq/day. The vast majority of intestinal K+ absorption occurs in the small intestine; the contribution of the normal colon to net K+ absorption and secretion is trivial. K+ is absorbed or secreted mainly by passive mechanisms; the rectum and perhaps the sigmoid colon have the capacity to actively secrete K+, but the quantitative and physiological significance of this active secretion is uncertain. Hyperaldosteronism increases fecal K+ excretion by about 3 mEq/day in people with otherwise normal intestinal tracts. Cation exchange resin by mouth can increase fecal K+ excretion to 40 mEq/day. The absorptive mechanisms of K+ are not disturbed by diarrhea per se, but fecal K+ losses are increased in diarrheal diseases by unabsorbed anions (which obligate K+), by electrochemical gradients secondary to active chloride secretion, and probably by secondary hyperaldosteronism. In diarrhea, total body K+ can be reduced by two mechanisms: loss of muscle mass because of malnutrition and reduced net absorption of K+; only the latter causes hypokalemia. Balance studies in patients with diarrhea are exceedingly rare, but available data emphasize an important role for dietary K+ intake, renal K+ excretion, and fecal K+ losses in determining whether or not a patient develops hypokalemia. The paradoxical negative K+ balance induced by ureterosigmoid anastomosis is described. The concept that fecal K+ excretion is markedly elevated in patients with uremia as an intestinal adaptation to prevent hyperkalemia is analyzed; we conclude that the data do not convincingly show the existence of a major intestinal adaptive response to chronic renal failure.