Thirst and hyperdipsia of anuric chronic uremics on maintenance hemodialysis and the possible dipsogenic factors were studied. Exaggerated thirst was present in 213 (86%) of the 247 studied patients. It usually started 4-6 h after the end of the dialysis session, persisted during the whole interdialytic period and often disappeared during the subsequent dialysis. Hyperdipsia, as indicated by the high body weight gain ( > 4%) in the interdialytic periods, was present in 33.6% of patients. The highest rate of increase of body weight occurred in the first hours following the end of dialysis sessions. Hypematremia, potassium depletion, increasing plasma urea levels and elevated plasma angiotensin II levels were considered as the possible dipsogenic factors of a nonpsychic nature. Sodium is certainly of paramount importance for its obliged extracellular position, and when sodium intake is elevated, hypematremia is very likely the cause of exaggerated thirst and weight gain in patients on hemodialysis. Potassium depletion may cause thirst in animals, but this condition is extremely rare in patients on maintenance hemodialysis, who often accumulate it. In these patients it is, therefore, unlikely that potassium depletion is a dipsogenic factor. Increasing serum urea levels exert an evident dipsogenic effect in anephric rats and urea, when infused into normal volunteers, stimulates thirst. The extracellular urea levels in the interdialytic period are certainly higher than the intracellular ones, as a consequence of its continuous accumulation, and this creates an osmotic gradient with a dipsogenic effect. When this gradient is reversed, following hemodialysis (which removes first the extracellular urea), the dipsogenic effect disappears. The hypothesis of a dipsogenic effect of urea operating in the interdialytic period in anuric patients on hemodialysis is therefore formulated. Angiotensin II is regarded as dipsogenic in patients on maintenance hemodialysis because of its high plasma concentrations. The following evidence is, however, against this contention: the ACE inhibitors do not prevent hyperdipsia, the body weight changes due to hyperdipsia are not correlated with the plasma levels of angiotensin II, and, finally, thirst is often absent in the hours of maximum angiotensin II plasma levels. In conclusion, hypematremia (frequently) and increasing plasma urea levels (regularly) appear to be the dipsogenic factors operating in patients on maintenance hemodialysis. The role of angiotensin II is doubtful and that of potassium depletion quite unlikely. Psychogenic factors may play an important role, however, in some patients.