Abstract Background Aging and obesity are two global concerns in public health. As obesity is associated with several metabolic disorders, it is a well-known risk factor of heart failure (HF). However, several studies have suggested that obesity may have paradoxical survival benefits in patients with HF. Therefore, the aim of this study was to investigate whether the obesity paradox exists in super elderly patients with HF in Japan. Methods A total of 4,292 patients from our hospital HF registry were divided into four age subgroups: (A) less than 65, (B) elderly 65-74, (C) high elderly 75-84, (D) super elderly ≥85. The primary endpoint was the all-cause death within 1-year after discharge. Results At baseline, the proportion of men in each group was 42-76%, the mean left ventricular ejection fraction was 49-61%, the average of N-terminal pro-B-type natriuretic peptide was 1972-2944 pg/ml, and the mean body mass index (BMI) ranged 21-25 kg/m2. Patients with higher BMI were more common in the younger age group, while those with lower BMI were more common in the older age group. Cubic spline curves of Cox proportional hazard regression models demonstrated an inverse relationship between BMI and mortality (Fig. 1). Lower BMI was associated with higher all-cause mortality in all ages, although this association was attenuated at the young age, and pre-obesity (25–29.9 kg/m2) was associated with the greatest survival benefit. Meanwhile, the impact of underweight (<18.5 kg/m2) was stronger super elderly subjects, a one-unit increase in BMI was associated with a 10.3% decrease in the hazard ratio of death (P=0.021). Kaplan-Meier curves for all-cause mortality also showed that lower BMI was associated with higher all-cause mortality in all groups, and the differences were statistically significant in Groups B, C, and D (Fig. 2). Conclusion Our study suggested the existence of the obesity paradox among HF patients in super elderly. Mildly overweight may be beneficial at old age with HF.