AimThis observational study aimed to examine the association between the A Body Shape Index (ABSI) and/or sarcopenia and total, cardiovascular, and cancer mortality. MethodsThe associations of sarcopenia and ABSI with all-cause, cardiovascular, and cancer mortality were assessed in 4,488 participants from the 1999-2004 National Health and Nutrition Examination Survey (NHANES) who were followed up until December 31, 2015. Models were analyzed separately for men and women and adjusted for age, race, and other confounding factors. ABSI was assessed as a continuous measurement by quartile for men and women. Population attributable fractions (PAFs) were calculated to assess mortality caused by sarcopenia and/or ABSI in the study population. ResultsWhen ABSI was assessed as a continuous variable, the ABSI quartile showed a linear trend for total (p = 0.0001), cardiovascular (p = 0.04), and cancer (p = 0.02) mortality in men and for total (p = 0.06) and cardiovascular (p = 0.06) mortality in women. The hazard ratios (HRs) of the fourth ABSI quartile were 1.51 [95% confidence interval (CI): 1.20-1.89] in men and 1.23 (95% CI: 0.93-1.64) in women, compared with those in the first quartile. When ABSI was assessed by quartile, the appendicular skeletal mass index (ASMI) was lower in the groups with high ABSI. When high ABSI was combined with sarcopenia, the HRs of all-cause mortality were 2.05 (95% CI: 1.60-2.62) in men and 1.51 (95% CI: 1.19-1.92) in women. In the subpopulation (sarcopenia group or higher ABSI), the PAFs of mortality due to sarcopenia were 26.16% (95% CI: 12.68-37.56) in men and 21.89% (95% CI: 5.64-35.35) in women, and the PAF of mortality due to higher ABSI was 23.70% (95% CI: 12.11-33.77) in men. ConclusionThe ABSI value was significantly associated with all-cause and cardiovascular mortality, and the co-existence of higher ABSI values and sarcopenia can contribute to a more significant death risk in comparison with high ABSI values or sarcopenia. Moreover, the ABSI values in combination with the ASMI can be used to preliminarily evaluate the content and distribution of fat and muscle and to predict the risk of death in obese and sarcopenic populations.