The yields of maize in Malawi are generally low under smallholder farmers. However this can be improved through introduction of legumes in the farming system. Grain legumes have been recognized worldwide as an alternative mean of improving soil fertility through their ability to fix atmospheric nitrogen, increase soil organic matter and improve general soil structure. The promotion of legume technologies needs to take into consideration other benefits, not just soil fertility improvement, if they are to be accepted. The land allocated to legumes remains low as compared to other crops (maize, tobacco) which are considered to be of social and economic importance. The fluctuation of legume prices especially for soybean and limited access to improved technologies of production such as use of inoculants and improved varieties contributes to low investment. The purpose of this study was to evaluate of the performance of legume. A comparison was made between the 2 districts of Salima and Mchinji on the legume trial packages especially soybean trial package to evaluate the technology suitability. Four types of legumes tested included: groundnut, soybean, bean and cowpea with different varieties. Planting was done between late December 2010 and late January 2011. Triple super phosphate (TSP) fertilizer was applied to the trial plots for all legumes whereas soybean was also inoculated with Marondera (Zimbabwe) or Chitedze (Malawi) inoculants. The extension agents involved in dissemination included Clinton Hunter Development Initiative (CHDI) and Department of Agriculture Extension services (DAES). The data were collected through interviews from famers participating in the project. A total 96 farmers were interviewed; 52 from Mchinji (Kalulu and Mkanda Extension Planning Area - EPA) and 44 from Salima (Chinguluwe EPA). Thirty farmer clubs participating in the project were targeted. The area under legume production was analyzed and expressed as proportion of total land being owned by individual farmers in the two districts. The yield estimates were determined based on the grain weight from the whole trial plot (20 x 10) m2 or net trial plot (16x6) m2. It was not possible to assess the total biomass because most famers did not weigh stovers hence the analysis is based on grain weight only. Access to extension services were assessed by interviewing farmers to inquire if both lead and secondary farmers had been visited by extension workers. The study results showed that the use of inoculants, fertilizer and right varieties improved legume production in smallholder farming systems. However the land allocated to legume production was small compared to the total available land. The implementation of large scale projects through partners is an effective way to reach more communities. However the assessment of partners should include availability of human capital at grassroots level. This contributed to differences in technology performance apart from biophysical and agronomic factors differing between Mchinji and Salima.