After observing alterations in the activity of the serum complement in patients undergoing periodic hemodialysis, the authors performed a preliminary study to determine repeatedly the activity of C3, C4, and CH50 in 44 patients. They discovered a consistent drop in C3 and CH50 while C4 remained normal. An attempt to explain these findings with information from the literature offered no more than a hypothesis for further study. While the possibility of a decline in the synthesis of the complement factors cannot be disregarded, the authors believe it is much more probable that they are consumed at a rate higher than normal. Since the C4 factor does not appear to be involved, activation is probably along the alternative pathway. Defective synthesis cannot be attributed to liver disease because the latter is not always present and because there is no relationship between C3 levels and levels of albumin or the presence of hepatopathy. The literature was reviewed for data that might support the idea that the complements in these patients are activated continuously by some process in connection with dialysis, by chemical products employed for to clean the machines, by commonly administered drugs, etc. Because so few data could be found on the subject, the authors consider that is necessary to study these mechanisms and their repercussions over a longer period of time.