beta-2 Microglobulin levels (beta-2M) were analyzed at four month intervals in sera of 237 patients on various forms of dialysis over a 2 year period; twelve patients volunteered to participate in short-term kinetics studies. Duplicate beta-2M measurements in biologic fluids were performed using an RIA kit. The data presented confirm elevated serum beta-2M in dialyzed patients whatever the dialysis method used and give an overview on various factors affecting circulating serum beta-2M. Intraindividual beta-2M variations were 13% in stable ESRD patients. Unusually high values (greater than 2 SE) were observed in patients presenting with severe intercurrent disease (e.g., cancer). The only significant difference observed between groups of patients, HD (46 +/- 1 mg/L) and PD (IPD: 33 +/- 3, CAPD: 37 +/- 2 mg/L), was due to the residual renal function preserved in a higher percentage of PD patients. No significant difference was noted in HF (40 +/- 2 mg/L) and HDF (38 +/- 5 mg/L) patients, despite a higher beta-2M removal rate. beta-2M membrane permeability differs greatly among filters. It is high for AN69 and Polysulfone, intermediate for Polyamide, low for Polyacrylonitrile and none for Cuprophane. Peritoneal membrane is highly permeable to beta-2M with beta-2M dialysate/serum levels of 0.20 at the end of a 6 hr exchange. In conclusion, beta-2M determination in uremic patients is useful in assessing material permeability and biocompatibility. However, it is too early to determine the impact or the beneficial effects of using highly permeable membrane to prevent beta-2M-amyloidosis related nonarticular complications.