ABSTRACT Introduction Galectin-3 plays critical roles in the adhesion, proliferation, and differentiation of tumor cells. Recent data have suggested that galectin-3 plays a role in the development of hepatocellular carcinoma (HCC); however, its prognostic value has not been validated. The aim of our study was to evaluate the clinical and prognostic value of galectin-3 in patients with HCC. Methods We prospectively enrolled and collected clinicopathologic data and serum samples from 767 patients with HCC between 2001 and 2014 at The University of Texas MD Anderson Cancer Center. Two hundred patients without HCC were also enrolled and had data collected. The Kaplan-Meier method was used to estimate overall survival (OS) distributions. Results The median OS in this cohort was 14.2 months (95% CI, 12–16.1). At the time of analysis, the 1-year OS rate was 45% (95% CI, 0.4–0.51) among patients with high galectin-3 levels and 59% (95% CI, 0.54–0.63) among patients with low galectin-3 levels. OS was significantly inferior in patients with high galectin-3 levels than in patients with lower galectin-3 levels (median OS: 10.12 vs. 16.49 months; p = 0.0022). Additionally, the multivariate model showed a significant association between high galectin-3 level and poor OS (hazard ratio [HR] = 1.249; 95% CI, 1.005–1.554). Comparison between low (n = 464 patients) and high (n = 302 patients) galectin-3 levels showed that mean serum galectin-3 levels were significantly higher in patients with HCC who had hepatitis C virus (HCV) infection (p = 0.0001), higher Child-Pugh score (CPS) (p = 0.0009), and higher Cancer of the Liver Italian Program (CLIP) score (p = 0.0015). Conclusion Our study shows that serum galectin-3 level is a valid prognostic biomarker candidate.