Background/Aim: The major limitation of carbohydrate antigen (CA)19-9 as a tumor marker is the high incidence of false-positive results during cholestasis. We evaluated preoperative CA19-9 and its adjusted values [ratios of CA19-9 to total-bilirubin (TB), direct-bilirubin (DB), and alkaline phosphatase (ALP)] to investigate the most suitable prognostic parameter in extrahepatic biliary tract cancer (eBTC) patients with or without jaundice. Patients and Methods: eBTC patients (n=140) who underwent resection were divided based on the absence (TB <2.0 mg/dl, n=90) or presence (TB ≥2.0 mg/dl, n=50) of preoperative jaundice. Within each group, the associations with overall survival (OS) were assessed for CA19-9, CA19-9/TB, CA19-9/DB and CA19-9/ALP ratios using Cox regression, receiver operating characteristic (ROC) analyses, and area under the curve (AUC) estimates. Results: In univariate analysis in the group without jaundice, both high CA19-9 and high CA19-9/TB ratio were associated with poor OS, whereas other parameters were not. ROC-AUC for OS prediction was greater in CA19-9 than in the CA19-9/TB ratio, and CA19-9 was identified as an independent prognosticator in multivariate analysis. In the group with jaundice, CA19-9 was not significant; however, CA19-9/TB, CA19-9/DB, and CA19-9/ALP ratios were all associated with poor OS. In ROC-AUC analysis, CA19-9/ALP ratio showed the highest predictive value; furthermore, it was an independent prognosticator in multivariate analysis. Conclusion: Adjustment of the CA19-9 value was less useful as a predictor in the absence of jaundice. On the other hand, the CA19-9/ALP ratio showed superior prognostic value in jaundiced patients with eBTC.