Experiments on deforestation and arable land development were established in 1978-79 at the research farm of the International Institute of Tropical Agriculture (IITA), Ibadan, Nigeria. The felled biomass in mechanized land clearing watersheds was collected in windrows constructed across the slope. These windrows were about 10 m wide and along the entire width of the watershed. W~ndrows were burnt in mid March 1979 about 8 weeks after felling the trees. Maize (Zea mays) and cowpea (Vigna unguiculata) were grown during the first (April to August) and second (September to November) seasons, respectively. Maizecowpea were double cropped each year. Considerable variability in growth and yield of maize and cowpea was observed seven to eight years after deforestation during 1985 and 1986. Based on crop vigor and color, 2 m X 2 m plots were categorized "good," "average" and "poor." Five microplots of each category were identified in each watershed. Grain yields were measured separately on these microplots. Grain yield was in the order of 4:2.1:1 for maize and 2.95:1.68:1 for cowpea for good, average and poor microplots respectively. Cowpea grain yield was ~ign~cantly correlated with ;oil organic C (R2 = 0.657) exchangeable Ca+2 (R~ = 0.621 , exchangeable Mg (It2 = 0.622; and cation exchange capacity (R2= 0.559). There was a negative correlation between cowpea grain yield and the MII+~ content (R2 = 0.346). Detailed soil evaluation revealed that all microplots with good crop growth were located within Ule former windrow zone. Crop growth and yields were signif~cantly superior in windrows than outside the windrow zone. Maize grain yield in the windrow zone was 22% and 267% more than innon-windrow zone in 1985 and 1986, respectively. Differences in crop yield were more shiking in seasons with poor rains (e.g., 1986) than those with good rains (e.g., 1985).