Background/Aims: In most cases of drug-induced liver injury, it is I difficult,6 diagnose whether these cases would,progress to fulminant hepatic. failure. We,investigated the characteristics I of non-viral and suspiciously drug-induced fulminant hepatic failure by comparing clinical data between cases that progressed and those that did: not progress to fulminant hepatic failure. Methodology: Ninety-five cases 4 suspicious drug-induced liver injury including 22 cases that had been treated at our hospital, and subsequently progressed to fulminant hepatic failure were involved in this study. We:investigated the characteristics of drug,induced fulminant hepatic failure by a comparison of non-fulminant and fulminant cases, and simultaneously of survivors and fatal cases in the group of fulminant cases. Results: Many of. the clinical variables were significantly deteriorated in fulminant cases. The latent period, which,means the,duration of drug administration, correlated with the severity of drug-induced liver injury including fulminant hepatic failure. Suspicious cases of drug-induced liver injury where, the bilirubin level at the, time of definite diagnosis stayed over 13mg/dL for more than one month were likely to progress to fulminant hepatic failure. Conclusions: Our results suggest that the latent period and the-peak level,of total bilirubin would be prognostic, factors for the severity,of drug-induced fulminant hepatic failure. Early preparation of liver transplantation should be recommended by referring these characteristics.