Worldwide, botanical gardens cultivate around 80,000 taxa, corresponding to approximately one-quarter of all vascular plants. Most cultivated taxa are, however, held in a small number of collections, and mostly only in small populations. Lack of genetic exchange and stochastic processes in small populations make them susceptible to detrimental genetic effects, which should be most severe in annual species, as sowing cycles are often short. In order to assess whether ex situ cultivation affects genetic diversity of annuals, five annual arable species with similar breeding systems were assessed with 42 in situ populations being compared to 20 ex situ populations using a random amplified polymorphic DNA (RAPD) analysis approach. Population sizes tended to be lower under ex situ cultivation and levels of genetic diversity also tended to be lower in four of the five species, with differences being significant in only two. Ex situ populations showed incomplete representation of alleles found in the wild. The duration of cultivation did not indicate any effect on genetic diversity. This implies that cultivation strategies resulted in different genetic structures in the garden populations. Although not unequivocally pronounced, differences nonetheless imply that conservation strategies in the involved gardens may need improvement. One option is cold storage of seeds, a practice that is not currently followed in the studied ex situ collections. This may reflect that the respective gardens focus on displaying living plant populations.