AbstractThe purpose of this study was to dissociate the authentic and artifactual effect ofin vivoheparin on drug protein binding using protamine as an inhibitor ofex vivolipolysis. A mixture of ethylenediamine tetra‐acetic acid (EDTA, 5 mg ml−1) and protamine in the concentration range of 0 to 7·5 mg ml−1was added to blood samples from 23 cardiac catheterized patients before (control) and 10 min after 3000 IU of intravenous heparin. In control samples, protamine does not interfere with the protein binding of lidocaine (L), quinidine (Q) or propranolol (P) when plasma pH is readjusted to 7·4. In the absence of protamine, heparin induced a significant increase in the free fraction by 40, 130, and 30 per cent for L, Q, and P, respectively (p<0·001), while free fatty acid (FFA) levels increased 2 to 6 fold. When protamine was present, the heparin‐induced elevation in free fraction was significantly lower for L (16 per cent) and Q (77 per cent) but not for P; FFA levels were decreased at all protamine concentrations. Residual increases in free fraction and FFA levels compared to control values may represent the truein vivoeffect of heparin at the peak activity of lipoprotein lipases. For L and Q, variations in free fraction were strongly associated with variations in FFA, but for P, no significant correlation was observed (r= 0·492). These results indicate that variations in free fraction of L and Q caused by heparin are, to a large extent, artifactual but may be prevented by use of protamine in collection tubes (5 to 7·5 mg ml−1). For P, the increase in free fraction was not mediated by variations of FFA indicating that another mechanism mu