The aim of this study was to assess the influence of host genetic factors on response to combination therapy for chronic hepatitis C infection.Patients with biopsy-proved chronic hepatitis C infection were treated with interferon alone (n = 143) or combined therapy of interferon + ribavarin (n = 105; 46 treatment naïve, 59 relapsers). Human leukocyte antigen (HLA) class I was determined by microlymphocytotoxicity and class II by polymerase chain reaction-single specific oligonucleotide. The two biallelic tumor necrosis factor-alpha promoter polymorphisms were studied by a polymerase chain reaction-amplification refractory mutation system. Other variables measured were viral genotype, hepatitis C virus RNA load, liver function tests, and ferritin concentration.Univariate analysis indicated that patients bearing HLA B44+, DRB1*03, infected by genotype non-1, with higher concentrations of transaminases and shorter duration of infection showed a higher sustained response (SR) rate than those on combination therapy. HLA class II and TNF-alpha promoter polymorphisms were not related to SR. In multivariate analysis, non-1 genotype (OR 2.42, 95% CI 1.12-5.55, p = 0.026) and HLA B44+ (OR 4.84, 95% CI 1.3-17.8, p = 0.017) were the independent variables associated with SR. However, HLA B44+ was not associated with SR in patients treated with interferon alone.HLA class I B44 is related to a higher rate of SR in combination therapy but not in interferon monotherapy, whereas HLA class II, tumor necrosis factor-alpha -238A or -308A seem not to influence response to the antiviral therapy. These findings may be of value in therapy selection for hepatitis C-infected patients.