The present study represents the follow-up of our initial observations designed to investigate whether in patients with nonalcoholic steatohepatitis (NASH) the beneficial effect of 12-week course of melatonin (MT) on liver enzymes could be maintained with prolonged period of treatment and to analyze whether biochemical treatment responses could be sustainable after melatonin discontinuation. Forty two patients with histologically proven NASH (30 treated with melatonin 2x5 mg daily, 12 controls receiving placebo) enrolled to our previous 3-month study agreed to take part of subsequent 12 weeks treatment followed by 12-week follow-up period. Enrolled patients had biochemical determinations every six weeks during the melatonin treatment period and again after 12 weeks of follow-up. Significant reduction in median alanine aminotransferase (ALT) levels between baseline and week 18, week 24 and follow-up was observed in both MT-treated and control group: 43% and 31%, 42% and 33%, 32% and 31%. Aspartate aminotransferase (AST) and gamma-glutamyl transpeptidase (GGT) levels decrease significantly only in MT-treated group. In MT-treated group mean percentage change in AST level below baseline at week 18, at week 24 and at follow-up was 45%, 33% (p<0.05) and 8% (ns), respectively. The evolution of GGT levels was as follows: the mean percentage reduction in GGT below baseline level at week 18, 24 and follow-up was: 48%, 52% and 38% (p<0.05), respectively. In both MT-treated and control group plasma cholesterol, triglicerydes and glucose concentrations as well as plasma alkaline phosphatase persisted within normal values during the prolonged study period. Plasma concentration of melatonin (pg/ml) in MT-treated group averaged 7.5±3.5 at baseline and increased to 52.5±17.5 at 24th week. The results of our study demonstrating beneficial effect of melatonin on liver enzymes in patients with NASH would seem to encourage further controlled trials of melatonin given over a longer period of time with liver histology as end point.