It is important to establish the various effects of legume cover crops on soil physicochemical properties because they have been considered for use as improved fallows (with shorter rest periods) to enhance development and maintenance of soil productivity. Our objectives were to assess: (i) aboveground dry matter yields of legume cover crops; and (ii) cover crop effects on weed infestation and soil physicochemical properties in a minimum tillage management system. Trials were conducted for 2 yr at Kawanda Agricultural Research Institute and on farmers’ fields in Mbale and Pallisa districts, eastern Uganda. Th e experiment layout was a Randomized Complete Block Design (RCBD) in a split-plot arrangement with four replications. Natural and improved fallows were established in the second cropping season of 2004. Cover crops used in the improved fallows included mucuna [Mucuna pruriens (L.) DC.var. utiliz], Dolichos lablab (Lablab vulgaris Savi cv. Rongai), canavalia [Canavalia ensiformis (L.) DC.], and crotalaria (Crotalaria paulina Schrank). The fallows were reestablished in the same plots in the second cropping season of 2005 aft er maize (Zea mays L.). Canavalia yielded significantly more dry matter than the other fallows regardless of year or site. With an average yield of 169 kg N ha–1 canavalia accumulated significantly more N than the other fallows; all improved fallows produced significantly more N than the natural fallow. Canavalia also accumulated significantly more P than the other fallows; all improved fallows, with the exception of crotalaria, accumulated more P than the natural fallow. There was no significant change in soil physicochemical properties by the improved fallows. All effects considered, improved fallows were more beneficial than natural fallow. A significant improvement in soil physicochemical properties using legume cover crops might be possible, though it may require more than the two cropping cycles used in this study of degraded soils.