Background <p>Anthropometric indicators reflecting overall body mass and fat distribution may be relevant to prognosis in patients with cancer. The weight-to-waist ratio (WWR), which integrates body weight and waist circumference, has recently emerged as a potentially informative marker of body composition status. However, its association with mortality after cancer diagnosis remains unclear.</p> Methods <p>This cohort study included participants who were cancer-free at baseline and subsequently developed incident cancer during follow-up. All anthropometric measurements, including WWR, were obtained at the 2006 baseline examination, prior to cancer diagnosis for all included participants. In survival analyses, time zero was defined as the recorded cancer-diagnosis time available in the HRS follow-up data, and only post-diagnosis follow-up time was included in the risk set. Because the interval between baseline assessment and cancer diagnosis varied across individuals, exposure misclassification may have occurred. In addition, because incident cancer in HRS may be identified between survey waves, some imprecision in the exact diagnosis date cannot be excluded. The primary outcome was all-cause mortality after cancer diagnosis. Cox proportional hazards models were used to evaluate the association between WWR and mortality with sequential adjustment for demographic variables, lifestyle factors, and comorbidities. WWR was analyzed as both a continuous variable and quartiles. Restricted cubic spline analyses, subgroup analyses, proportional hazards assumption testing, and sensitivity analyses were also performed.</p> Results <p>A total of 834 participants with incident cancer were included, among whom 331 deaths occurred during follow-up. The median age was 66.0&#xa0;years (interquartile range [IQR], 59.0–72.0), and 426 participants (51.1%) were male. In the fully adjusted model, each 1-standard deviation increase in WWR was associated with a lower risk of all-cause mortality (hazard ratio [HR] = 0.76, 95% confidence interval [CI] 0.66–0.87; <i>P</i> &lt; 0.001). When modeled continuously per 1-unit increase, the corresponding HR was 0.06 (95% CI 0.02–0.24; <i>P</i> &lt; 0.001), but SD-standardized estimates were considered more interpretable because WWR had a relatively narrow range. Compared with the lowest quartile, the adjusted HRs were 0.70 (95% CI 0.52–0.95) for Q2, 0.57 (95% CI 0.41–0.80) for Q3, and 0.57 (95% CI 0.39–0.83) for Q4. Restricted cubic spline analysis showed a significant overall association (<i>P</i> &lt; 0.001) without strong evidence of nonlinearity (P for nonlinearity = 0.077). WWR showed slightly better discrimination (C-index 0.757) than BMI (0.749) and WC (0.749). The proportional hazards assumption was satisfied (global Schoenfeld test, <i>P</i> = 0.623), and results remained materially unchanged after excluding participants who died within 2&#xa0;years after cancer diagnosis.</p> Conclusions <p>Higher baseline pre-diagnosis WWR was associated with lower all-cause mortality after cancer diagnosis in this cohort. WWR may be associated with prognosis, but this observational finding should be interpreted cautiously. Further validation in cancer-specific cohorts with detailed information on tumor type, stage, treatment, and repeated anthropometric measurements is needed before clinical application.</p>