Summary. HCV genotypes 2- or 3-infected patients with a rapid virological response (RVR) to therapy with pegylated interferon and ribavirins who have a low viral load, noncirrhotic and nonobese may be considered for a shorter course of treatment. However, no studies have assessed host-viral factors associated with relapse in genotype 2 and 3 separately. Accordingly, we assessed whether 12 weeks of pegylated interferon and ribavirin was an optimized regimen for treatment of HCV genotype 2 and 3 with positive predictors of response. Power and sample size were a priori calculated and 96 consecutive chronic hepatitis C patients (53, genotype 2 and 43, genotype 3) without cirrhosis who were not obese and who achieved a RVR to therapy with peg-IFN-α-2a and ribavirin were enrolled. Fibrosis, steatosis, homeostatic model assessment-insulin resistance and HCV RNA were predefined variables to be evaluated in relapse. An intention-to-treat analysis was performed. SVR rates were 98% and 84% for genotype 2 and 3, respectively. Analysis of genotype 3 patients who had relapse showed a negative correlation with steatosis (P < 0.0001) and HCV RNA (P < 0.015). Multivariate analysis showed that steatosis was the independent predictor of relapse (OR, 0.988; 95% CI, 0.981–0.993; P < 0.001). Genotype 3 patients with steatosis had a relapse rate of 36.4% and 15.8% in those with high and low viral load, respectively, whereas there was no relapse in those without steatosis. In conclusion, a 12-week course of therapy is sufficient for patients without cirrhosis, not obese and infected with HCV genotype 2 achieve a RVR. This is not the case for genotype 3. Steatosis is the independent predictor of relapse. New therapeutic strategies are necessary for this subgroup of HCV genotype 3.