Within a population, individuals may differ in their demographic traits – their propensity to survive or their expected reproductive success. When linked to age or size, these differences can be incorporated into age- or size-structured models to project the population dynamics. However, differences may still remain after accounting for age and size – a phenomenon we call “demographic heterogeneity.” Prior work has demonstrated that demographic heterogeneity may influence the variance in population growth rate due to demographic stochasticity, and hence influences the extinction risk of small populations. In that work an individual’s traits were uncorrelated across years. We now extend these models to consider situations where an individual’s demographic traits – e.g., greater than average or lesser than average survival probability – are acquired at birth and retained throughout the individual’s life. Using branching process models to calculate extinction risk, we show that in a non-age-structured population, heterogeneity in individual survival probabilities decreases extinction risk, whereas heterogeneity in fecundity slightly increases risk. In addition to influencing the demographic variance, heterogeneity may also influence the mean and variance of the distribution of trait values.