The recent reports of a very high frequency of signs of hepatitis C virus (HCV) infection among patients with essential mixed cryoglobulins (EMC) suggest new hypotheses for the pathogenesis of this disease. However, most of these studies have been seriously criticized. The serologic test designed for detection of anti-HCV antibodies (ELISA, RIBA I and II) may yield a significant rate of false-positive results when performed on cryoglobulinemic sera, and the detection of the HCV genome by PCR is still heavily conditioned by practical problems. Indirect, but possibly more reliable, evidence of HCV infection in cryoglobulinemic patients might come from the demonstration of anti-HCV antibodies by a conventional technique (ELISA or RIBA) in the purified polyclonal non-rheumatoid immunoglobulinemic fraction excreted in the urine by glomerular filtration. Fifty-two patients whose serum had tested positive for HCV antibodies (by ELISA and RIBA) on multiple occasions were enrolled in this study. They were diagnosed as having either EMC or HCV chronic hepatitis without cryoglobulinemia at least one year ago. The urine samples of these patients were tested for anti-HCV antibodies by ELISA and RIBA. In patients with chronic C hepatitis the antibodies most frequently found in the serum were anti-C33c and anti-C22-3. The results of the RIBA were substantially confirmed by ELISA, with a positive test in the urine of 30 of 32 seropositive patients. Similar results were obtained in patients with EMC II. We conclude that the specificity of the RIBA and ELISA tests for HCV antibodies in patients with EMC appears to be as high as in HCV+ patients without serum cryoglobulins. EMC patients have a high incidence of HCV infection and active chronic liver disease.