Theory predicts that populations of animals with late maturity, low fecundity, large body size and low body growth rates will have low potential rates of population increase at low abundance. If this is true, then these traits may be used to predict the intrinsic rate of increase for species or populations, as well as extinction risks. We used life–history and population data for 63 stocks of commercially exploited fish species from the northeast Atlantic to test relationships between life–history parameters and the rate of population increase at low abundance. We used cross–taxonomic analyses among stocks and among species, and analyses that accounted for phylogenetic relationships. These analyses confirmed that large–bodied, slow–growing stocks and species had significantly lower rates of recruitment and adult production per spawning adult at low abundance. Furthermore, high ages at maturity were significantly correlated with low maximum recruit production. Contrary to expectation, fecundity was significantly negatively related to recruit production, due to its positive relationship with maximum body size. Our results support theoretical predictions, and suggest that a simply measured life–history parameter can provide a useful tool for predicting rates of recovery from low population abundance.