Objective To verify the ability of nutritional indicators to predict risk of hospitalization in hemodialysis patients. Methods This prospective cohort study was conducted in two hemodialysis units in Southern Brazil. The following nutritional indicators were evaluated: subjective global assessment, malnutrition-inflammation score, nutritional risk screening 2002, percentage of body fat, mid-upper arm muscle circumference, hand grip strength, phase angle, and serum albumin. The association between nutritional indicators and risk of hospitalization over a period of two years was analyzed. Results One hundred thirty-eight patients (55.4±15.2 years, 61.6% men) were evaluated. The cumulative incidence of hospitalization during the study was 48.0% (95%CI=37.9-58.0). Patients hospitalized during the study period had higher body mass index and lower serum albumin than those who were not hospitalized. Based on the other indicators, malnourished patients did not have a significant risk of hospitalization (p>0.05). Serum albumin was the only predictor of hospitalization. Patients with levels <3.8 g/dL had a density ratio for incidence of hospitalization 2.47 times greater than those with higher levels (p=0.003). Gender was a modifier of the effect (p interaction=0.042), with an effect in women of 7.31 (95%CI=2.34-22.9; p=0.001) and in men of 1.37 (95%CI=0.60-3.12; p=0.448). Conclusion Serum albumin was able to predict the risk of hospitalization over a period of two years in female patients undergoing hemodialysis.