CVD prediction models aim to estimate an individual's risk of developing CVD to guide prevention and screening. Many models have been developed, but their applicability across populations and age groups is unclear. This review examines the influence of age on the performance of CVD risk prediction models. Age is a strong, independent risk factor for CVD and is included in most prediction models. However, age thresholds for CVD risk categorization vary between models. For example, the Framingham risk score considers ages 55-65 years as a risk factor but SCORE considers ages 60-65 years. These differences influence risk classification, especially in older adults. Although age is an important predictor, its perceived influence depends on the outcomes and time horizons considered in each model. Models predicting hard CVD outcomes (e.g. MI) over shorter timeframes tend to more heavily weight age. Models considering softer outcomes (e.g. CVD mortality) over longer time horizons put less emphasis on age, as other factors like blood pressure and cholesterol levels become more predictive. The discrimination and calibration of models tend to decrease with advancing age, suggesting reduced accuracy in older populations. Models developed in younger populations often overestimate risk in older adults. Some models have been adapted or re-calibrated for use in older adults, with moderate success. However, few models have been validated in adults over 75 years - the fastest growing demographic group. In summary, age is an important but imperfect predictor of CVD risk.The perceived influence and accuracy of age as a predictor depends on the outcomes and populations that models are developed and applied in. CVD prediction models have limited validation and applicability in older adults, particularly those over 75 years of age. Further research is needed to develop and validate models for this age group.