HEMODIALYSIS is remarkably safe. Death due to technical failure or humanerror occurs in only about 1 in 75,000 treatments. The resilience of the cardiovascular system can be appreciated from its ability to tolerate hours of extracorporeal blood circulation at flow rates of 300 mVmin or greater. Very few patients develop true shock during or as an immediate consequence of hemodialysis. A substantial proportion, however, become hypotensive to the extent that postdialysis malaise limits prompt return to predialysis physical activities in many otherwise well rehabilitated patients. Little attention was devoted to the understanding of dialysis-induced hypotension in the first decade of maintenance hemodialysis perhaps, because patient and physician viewed this complication as an unavoidable side effect of a wondrous new therapy. Subsequently, lessons learned in the mid 1970s, from clinical trials of postdilutional hemofiltration, and reexploration of the utility of bicarbonate (rather than acetate) buffering of dialysate, now suggest that dialysis-induced hypotension may be minimized or prevented under certain conditions. Keshaviah and Shapiro examine eight variables (their Table 1) which could account for a fall in blood pressure during or after dialysis and state that its genesis is . . complex, controversial, and multifactorial. With laudable caution, Keshaviah and Shapiro decline to indict a single causative mechanism though they conclude that at low dialysate sodium concentrations acetate . . is associated with a more marked fall of blood pressure than is bicarbonate. Heneghan accepts the evidence tying acetate to hypotension and explores the question of Why is acetate associated with greater morbidity during dialysis? In a simple but telling double-blind experiment, Heneghan administered oxygen or compressed air to five patients during 30 hemodialyses to test his thesis that acetate-caused hypoxemia was responsible for hypotension. In this study, when hypoxemia was prevented by oxygen, hypotension was avoided and dialysis was better tolerated. According to Heneghan, acetate metabolism reduces CO2 production which in turn induces hypoventilation and hypoxemia. Recognizing that long term comparisons of acetate and bicarbonate have not been done, Heneghan nevertheless concludes that . . acetate must be considered suspect. Directors of dialysis units must weigh the data for themselves to determme whether or not to invest in machines equipped to infuse bicarbonate as the principal dialysate buffer. In my view, the case is still open. Adding another pump increases the probability of mechanical error. Until the controlled studies called for by both essayists are performed, nearly all patients prone to hypotension can be managed by using a normal or high dialysate sodium concentration combined with administration of nasal oxygen throughout dialysis.