Gene flow in crop–wild complexes between phenotypically differentiated ancestors may transfer adaptive genetic variation that alters the fecundity and, potentially, the population growth (λ) of weeds. We created biotypes with potentially invasive traits, early flowering or long leaves, in wild radish (Raphanus raphanistrum) and F5 crop–wild hybrid (R. sativus × R. raphanistrum) backgrounds and compared them to randomly mated populations, to provide the first experimental estimate of long‐term fitness consequences of weedy life‐history variation. Using a life table response experiment design, we modeled λ of experimental, field populations in Pellston, MI, and assessed the relative success of alternative weed strategies and the contributions of individual vital rates (germination, survival, seed production) to differences in λ among experimental populations. Growth rates (λ) were most influenced by seed production, a trait altered by hybridization and selection, compared to other vital rates. More seeds were produced by wild than hybrid populations and by long‐leafed than early‐flowering lineages. Although we did not detect a biotype by selection treatment effect on lambda, lineages also exhibited contrasting germination and survival strategies. Identifying life‐history traits affecting population growth contributes to our understanding of which portions of the crop genome are most likely to introgress into weed populations.