The use of low-sodium diets in dogs with heart failure is common practice, but randomized, double-blind studies have not been conducted to examine the benefits or problems with this approach. The purpose of this study was to determine the effects of a low-sodium diet on clinical, echocardiographic, and neurohormonal parameters in dogs with heart failure. Dogs with stable chronic heart failure were fed exclusively a low-sodium (LS) and a moderate-sodium (MS) diet for 4 weeks each in a randomized, double-blind, crossover design. At days 0, 28, and 56, echocardiography and thoracic radiography were performed, and blood was analyzed for electrolytes and neurohormones. Fourteen dogs completed the study (9 with chronic valvular disease and 5 with dilated cardiomyopathy). Electrolyte abnormalities were common during the study, and serum sodium and chloride concentrations decreased significantly on the LS diet. Neurohormones did not change significantly between diet groups. Maximum left atrial (P = .05) and standard left atrial (P = .09) size decreased on the LS diet. For dogs with chronic valvular disease, vertebral heart score (P = .05), left ventricular internal dimension in diastole (P = .006) and systole (P = .02), standard left atrial dimension (P = .03), maximum left atrial dimension (P = .02), end-diastolic volume index (P = .02), and end-systolic volume index (P = .04) decreased significantly on the LS diet compared to the MS diet. Although analysis of these data suggests some benefits of a low-sodium diet, future studies with improved study design are needed to further evaluate the advantages and disadvantages of sodium restriction in dogs with heart failure.