Alpha interferon induces HBeAg seroconversion in about one third of treated patients and has become an established treatment of chronic hepatitis B. A number of smaller studies have suggested that response to treatment is more likely to occur in patients with higher levels of transaminases, with recent (adult) onset, a history of acute hepatitis, low levels of HBV DNA and in heterosexual males. The aim of this European co-operative study was to estimate the effect of alpha interferon more accurately and to evaluate the influence of host pre-treatment variables on the effect of interferon. Individual data were collected from 751 patients from 10 controlled clinical trials on alpha interferon (lymphoblastoid or recombinant) treatment for chronic hepatitis B. Alpha interferon was administered to 496 patients, while 255 were untreated controls. Individual patient data were analysed by survival analysis (log rank test and Cox regression analysis), stratified by trial, with the disappearance of HBeAg as the major endpoint. The results showed that the HBeAg disappearance rate with or without interferon treatment was higher in patients with high aminotransferase levels, with a history of acute hepatitis and in male heterosexual patients disregarding HIV status. If HIV-positive patients were excluded, the effect of sexual orientation was not significant. Therapy with alpha interferon increased the a priori HBeAg disappearance rate by a factor of 1.76; the relative treatment effect of alpha interferon was independent of the tested pretreatment host variables, but dependent on the total (intended) interferon dose (low dose < or = 200 MU/m2 increased HBeAg disappearance by a factor 1.37; medium/high dose > or = 200 MU/m2 increased HBeAg disappearance by a factor 2.05). In conclusion, this meta-analysis suggests that the effect of alpha interferon is less than previously assumed and independent of pretreatment host variables tested. It confirms the higher therapeutic benefit of a total dose exceeding 200 MU/m2 and of selection of patients based on disease activity and immune reactivity. Although all patient seem to have the same relative benefit, the absolute benefit of alpha interferon treatment seems to be greatest in patients with high transaminase levels and with a history of acute hepatitis.