In traditional yam (Dioscorea spp.) production systems in West Africa, finding sole seed yam producers is rare and up to 30% of harvested tubers from the ware tuber crop is reserved to plant an equivalent area of the harvested crop during the next season. Many farmers typically must make a trade-off between food and seed in the use of available tubers. A study was carried out using a factorial experiment in a randomised complete block design to investigate the influence of planting different minisett sizes at different planting periods on the yield of seed yam with the aim of improving the availability of this expensive input and saving more ware tubers for food or sale. Results showed that by doubling or tripling the minisett size from 30 to 60 or 90  g, yields of seed yam increased by 61.1 and 103.3%, respectively. The 90 g minisetts had the highest values for all traits studied but had the least sett multiplication ratio. The early planted crop yielded 137% more than the late-planted crop. Significant interactions between minisett size, planting period and season were found. In 2016, although the highest yield of seed tubers was from early planted 90  g minisetts (35.6 t ha−1), the yield of early planted 30 g minisetts (23.8 t ha−1) was similar to those of 60 g planted mid-season (28.7 t ha−1) and 90 g planted late (20.0 t ha−1). To produce a high proportion of seed size tubers with less planting material, early planting of 30  g minisetts is recommended. Such practice will enhance seed tuber availability and food security by saving 1–2 t ha−1 of tubers that would have been used as seed instead of food.