Rapid decline of the residual effect and ineffective weed suppression have been identified as the shortcomings for large-scale adoption of Sesbania sesban and Tephrosia vogelli planted fallows. An experiment was conducted at Msekera research station, Chipata in eastern Zambia from November 1998 to June 2002 on a Dystric Regosol to determine the residual effetcs of mixed fallows of trees combined with herbaceous legumes either archer dolichos (Macrotyloma axillare) or velvet bean (mucuna pruriens) compared with monospecies tree fallows on maize grain yields. Sole Gliridicia sepium fallow had the highest tree survival of 97% compared with the lowest 36% for Tephrosia vogelli + archer dolichos. At the time of clearing fallows two years after planting, Gliridicia + archer dolichos mixture recorded the highest surface litter of 3.49t ha~1 in contrast to Sesbania + Gliridicia and 4.50t ha~1 by Tephrosia + archer. Among the planted, fallows, Sesbania + archer mixture produced the greated maize yield of 2.7 and 1.9 t ha~1 in 2001 and 2002 seasons, respectively. These yields were, however, much lower than those produced by fertilized maize. Although planted fallows significantly increased maize yields compared with unfertilized maize monocrop, only sesbania + archer mixture had resulted in significantly higher yield over natural fallow. The planted fallows did not show significant residual effects compared with natural fallow beyond the first post-fallow maize. Both dolichos and velvet bean chocked the trees and there is a need to explore for herbaceous legumes that can yield some economic products and no doubt compete with the trees.