Experiments were conducted in Nigeria to evaluate the optimum plant population density for okra (Abelmoschus esculentus) sown in monoculture or intercropped with cassava (Manilhot esculenta). The crops were sown as late and early season crops in 1997 and 1998. Okra was sown at varied spacings to achieve final plant densities of 25000, 35000 and 50000 plants/ha in both monoculture and mixed stands. Intercropping had no significant effect on okra phenology (time to vegetative growth, flowering and fruiting). However, it significantly reduced weed growth by 25-45 % and nematode infection in okra by 36 %, and kept the canopy environment of cassava cooler by 3.0-4.4 degreesC and more moist by 5-10 g/kg, compared with monoculture. The durations of vegetative and reproductive growth and the weed-control ability of okra in both cropping systems were influenced by population density. Okra sown at 50000 plants/ha took the longest time to reach specific phenological stages, controlled weeds best and gave the highest fresh pod yields in both seasons. There was little increase in pod yields between 35000 and 50000 plants/ha in mixed stands in late season sown crops. The growth and tuber yield of cassava were not significantly affected by intercropping and okra population density. The growth of okra and cassava were better with early than late season sowing, irrespective of okra population density or cropping system. This appears to have been due mainly to the effect of temperature and rainfall differences between the seasons. The substantial variation in the weather during the two seasons also had greater effect in monoculture than mixed stands. It is concluded that when rainfall is limiting, okra could be intercropped with cassava using population density up to 35 000 plants/ha to allow the vegetable to develop and to maximize pod yield under the relatively dry conditions. The optimum target suggested is 50000-60000 plants/ha during the wet periods.