As part of conservation of plant genetic resources, long-term storage of seeds is highly relevant for genebanks. Here we present a systematic review and a meta-analysis of studies on seed longevity focusing on half-life (P₅₀) under different storage conditions. Six studies were selected for the meta-analysis; in addition, a high number of additional references were included in the discussion of the results. The results show that under ambient conditions, half-life is short, from 5 to 10 years, while under more optimal conditions, which for orthodox seeds is at low humidity and low temperature, half-life is more in the 40−60 years range, although with large interspecies variation. Under long-term genebank conditions, with seeds dried to equilibrium and thereafter kept at minus 18−20°C in waterproof bags or jars, half-life can be twice or three times as long. In general, many of the grain legume seeds, as well as corn, common oat, and common barley are long-lived, while cereal rye, onion, garden lettuce, pepper, and some of the forage grasses are more short-lived. Conditions during maturation and harvesting influence longevity, and proper maturation and gentle handling are known to be of importance. Seed longevity models have been developed to predict final germination based on initial viability, temperature, humidity, storage time, and species information. We compared predicted germination to results from the long-term experiments. The predicted values were higher or much higher than the observed values, which demonstrate that something in the seed handling in the genebanks have not been optimal. Long-term studies are now available with data at least up to 60 years of storage. Our review shows that the knowledge and methodology developed for the conservation of plant genetic resources should also work for wild species of orthodox seed nature.