Approximately half of all patients with chronic hepatitis C show an initial biochemical response to interferon, but only 15% to 20% of patients achieve a sustained response. We studied the efficacy of retreatment with interferon for patients with chronic hepatitis C who showed transient biochemical responses to initial treatment. Thirty patients who relapsed were retreated 1 to 52 months (median 14) after the end of initial treatment, according to the previously used regimens. The responses were correlated with the pre-retreatment patient data. The liver histologic grades, compared with those found before the initial treatment, were better in eight (27%) patients but worse in six (20%), whereas the fibrosis stage was improved in five (17%) but worsened in eight (27%). All patients displayed end-of-retreatment biochemical responses. Of the 30 patients, 10 (33%) achieved sustained aminotransferase normalization and serum hepatitis C virus (HCV) RNA clearance, but the remaining 20 patients showed relapse within 1 year after cessation of retreatment. Univariate analysis associated the sustained response with low pre-retreatment viral loads (0.8 +/- 0.7 MEq/mL vs. 9.1 +/- 6.5 MEq/mL; p = 0.006), short treatment intervals (13 +/- 13 months vs. 22 +/- 14 months; p = 0.031), and low histologic grades (1.3 +/- 0.7 vs. 1.9 +/- 0.7; p = 0.039). However, multivariate analysis indicated that only the pre-retreatment viral load was predictive of the sustained response (p = 0.049). These findings suggest that transient responders to interferon are likely to respond to retreatment but the achievement of a sustained response depends on the HCV viral load before retreatment.