Abstract Background and Aims Myosteatosis represents the infiltration of fat into the skeletal muscle in the early stages of sarcopenia, and has been suggested to be associated with adverse outcomes in patients undergoing dialysis. Skeletal muscle density (SMD) determined by computed tomography (CT) is commonly used to assess myosteatosis, but SMD is affected by contrast resulting in inaccurate measurements. Recently, indices for muscle quality measurements determined by muscle area have been developed. Thus, We aimed to investigate the relationship between a novel muscle area index and mortality in initial-dialysis patients and analyse variables related to myosteatosis using the cutoff values. Method This 3-year multicentric retrospective study enrolled a total of 782 initial-dialysis patients from four centres between 2014 and 2019 in China. SMD skeletal muscular area (SMA) normal attenuation muscle area (NAMA) intermuscular adipose tissue area (IMAT) was measured and skeletal muscle index (SMI) total attenuation muscle area (TAMA) NAMA/TAMA index was calculated on CT images at the level of the first lumbar vertebra. The relationship between NAMA/TAMA index and all-cause and cardiac death outcomes was analyzed. And the sex-specific optimal cut-off values of L1-NAMA/TAMA index were determined in relation to all-cause mortality. Univariate and multivariate Cox regression models were applied to evaluate potential risk factors for death. Results A total of 782 patients were enrolled. The optimal cut-off values of NAMA/TAMA index were 48.76 for males and 58.02 for females. Pearson's correlation coefficient between L1-NAMA/TAMA index and L1 SMD was 0.589 (P < 0.001), no significant association between L1-NAMA/TAMA index and L1 SMI was observed (P > 0.05), similarly, no significant association between L1 SMD and L1 SMI (P > 0.05). Multivariate regression analysis revealed that the low L1-NAMA/TAMA index group had lower risks of all-cause death (hazard ratio 0.988; 95% confidence interval 0.977–0.999, P = 0.041) and cardiac death (hazard ratio 3.74; 95% confidence interval 1.43–9.79, P = 0.007). Conclusion Our findings demonstrate that L1-NAMA/TAMA index acquired via CT scan is of great prognostic value on all-cause and cardiac death in patients undergoing initial-dialysis.