1 Plasma protein binding of disopyramide was determined twice in plasma from seven patients with severe nephrotic syndrome when the disease activity was markedly different for each patient (mean ± s.d. of serum albumin 21 ± 4 vs 29 ± 3 g 1-1 (P < 0.05), proteinuria 13.6 ± 9.6 vs 2.2 ± 1.7 g day-' (P < 0.01), and oal-acid glycoprotein 0.34 ± 0.12 vs 0.95 ± 0.28 g 1-1 (P < 0.01) during the exacerbation vs remission phases, respectively).2 Plasma samples containing disopyramide at the concentrations of 0.2-12.0 P,g ml-'were analysed by ultrafiltration. The free fractions at the proposed therapeutic concentration range (2.0-6.0 ,ug ml-l) were significantly (P < 0.01) greater during the exacerbation phase than during the remission phase. 3 Multiple linear regression analysis revealed that the free fraction of disopyramide at 3.0 ,ug ml-1 correlated much better with oal-acid glycoprotein (partial correlation coefficient = 0.85, P < 0.01) than with albumin (partial correlation coefficient = 0.25, P < 0.05). 4 The binding data of disopyramide analysed by a model assuming one specific binding site and nonspecific binding(s) demonstrated that the capacity constant correlated significantly (r = 0.97, P < 0.001) with plasma aot-acid glycoprotein. 5 The results suggest that a total concentration of disopyramide within the therapeutic range may not be a reliable guide for a safe dosing scheme in patients with severe nephrotic syndrome, particularly during the exacerbation period. It is recommended that free fraction of disopyramide and/or ctl-acid glycoprotein concentration should be frequently monitored in those patients with the changing disease activity to whom this antiarrhythmic is to be prescribed.