BACKGROUND/AIMS: Carbohydrate antigen 19-9 (CA19-9) is used as a biomarker to differentiate benign from malignant gastrointestinal disorders. We examined the value of CA19-9 measurement in polycystic livers after observing high CA19-9 cyst fluid levels in a benign polycystic liver case. METHODS: We determined CA19-9 levels in serum (n=120) and hepatic cyst fluid (n=81), from patients with polycystic livers (n=109) and simple hepatic cysts (n=24). Further, we analysed CA19-9 expression in normal and polycystic liver tissue (n=17). RESULTS: Cyst fluid CA19-9 levels from both polycystic livers and simple hepatic cysts were extremely high (median 91 000 U/ml, range 14-15 870 000 U/ml; median 85 000 U/ml, range 332-1 744 000 U/ml respectively). Serum CA19-9 levels were significantly higher in polycystic liver patients (median 30 U/ml, range 0-1200 U/ml) compared with patients with simple hepatic cysts (median 10 U/ml, range 3-200 U/ml, P=0.0011). Serum CA19-9 levels correlated with those in cyst fluid (r=0.3979, P=0.0399), polycystic liver volume (r=0.3870, P=0.0025) and the size of the largest cyst (simple cysts group; r=0.5319, P=0.0280). Cyst epithelia showed strong CA19-9 expression. Evacuation of cyst fluid in four patients resulted in a dramatic decrease in the serum CA19-9 levels (60-95%). CONCLUSIONS: CA19-9 levels are high in the cyst fluid and serum of polycystic liver disease patients due to production and secretion by cyst epithelia. It does not reflect malignancy in these patients and may be of value as a biomarker for intervention efficiency assessment.