A field experiment involving two plant densities (83,333 and 166,666 plants per hectare), two cropping systems (monoculture and mixed culture) and five cowpea (Vigna unguiculata L. Walp.) genotypes (3 farmer-selected varieties: Bensogla, Sanzie and Omondaw, and 2 breeder-improved cultivars: ITH98-46 and TVuI509) was conducted for two years in 2005 and 2006 at Nietvoorbij (33°54S, 18°14E), Stellenbosch, South Africa, to evaluate the effect of these treatments on the growth and symbiotic performance of cowpea. The results showed that, of the five cowpea genotypes, plant growth and N2 fixation were significantly greater in the three farmer-selected varieties (Sanzie, Bensogla and Omondaw) relative to the two improved cultivars (ITH98-46 and TVuI509). Furthermore, plant growth and symbiotic performance (measured as tissue N concentration, plant N content,15N natural abundance and N-fixed) were significantly (P<-50.05) decreased by both high plant density and mixed culture (intercropping). However, the %Ndfa values were significantly (P<-50.05) increased by both high plant density and mixed culture compared to low plant density or monoculture (or monocropping). Whether under low or high plant density, the cv. Sanzie was found to accumulate significantly greater total N per plant in both 2005 and 2006, followed by the other two farmer varieties relative to the improved cultivars. Similarly, the actual amount of N-fixed was much greater in cv. Sanzie, followed by the other farmer varieties, under both low and high plant density. The data also showed better growth and greater symbiotic N yield in cowpea plants cultivated in monoculture (or low plant density) relative to those in mixed culture (or high plant density). Our data suggest that optimising legume density in cropping systems could potentially increase N2 fixation in cowpea, and significantly contribute to the N economy of agricultural soils in Africa.