Background: Cancer survivors are at risk of higher mortality from other causes, in particular, cardiovascular and metabolic. We have previously reported on a development of a brief screening instrument to identify cardiometabolic predictors of mortality in older individuals using the Australian Longitudinal Study of Ageing, which showed that age, sex, history of cerebrovascular disease, non-adherence to exercise guidelines, and smoking predicted for mortality(Lim et al. 2017). The objective of this study was to validate our findings using an external database. Methods: Two external validation cohorts of cancer survivors recruited to the Australian Longitudinal Study on Women’s Health study were used: an older group aged between 70 and 75 at time of recruitment (cohort 1) and a younger group aged between 45 and 50 at time of recruitment (cohort 2). These were compared to the original development cohort who were aged 70 years and over at time of recruitment. The Cox proportional hazards model previously estimated in the development cohort was used to predict mortality at 10 years in the validation cohorts. Measures of discrimination and calibration were calculated. Results: There were 1764 and 1833 women in the two validation cohorts and 294 men and women in the original development cohort. Validation cohorts had lower mortality rates (31% and 4% in validation cohorts 1 and 2, respectively, versus 61% in the development cohort) and validation cohort 2 had a higher proportion of individuals reported to have met exercise guidelines (59%) compared to the development cohort (24%) and validation cohort 1 (0.11%). All measures of discrimination for the validation cohorts were within the range of acceptable published estimates. Calibration showed that, overall, the tool had acceptable predictive accuracy but performed best in the good prognostic groups. Conclusions: We have demonstrated that the screening tool developed in older cancer survivors, performs adequately in cohort of younger women cancer survivors although further calibration is needed to improve performance. Further validation in mixed gender cohorts and usability testing are planned.