Several agro-ecological practices have been developed to increase the range of soil fertility management options available to farmers, but little information is available on their labour input requirement and profitability relative to conventional practices. This study used a multiple visit cost-route method to estimate labour inputs and financial profitability of agroforestry-based and conventional (with and without fertiliser) soil fertility management practices in eastern Zambia. Results show that aggregated over a five-year cycle, labour inputs used under agroforestry practices were lower than fertilised maize fields but higher than non-fertilised continuous cropped maize. Agroforestry practices were more profitable (NPV ranging between $233 and $327 ha-1) than unfertilised maize fields (NPV of $130 ha-1) but less profitable than fertilised maize fields. The NPV for the latter is $499 ha-1 under a 50% government subsidy on fertiliser but, it fell to $349 ha-1 when the subsidy effect is removed. In terms of return per unit investment cost, agroforestry practices performed better with a benefit to cost ratio ranging between 2.77 to 3.13 in contrast to 2.65 (subsidised fertiliser) and 1.77 (non-subsidised fertiliser). The returns to labour per person day was $2.56 in (non-subsidised) fertilised maize field, $3.16 in subsidised fertiliser field, $2.63 in Gliricidia fields, $2.41 in Sesbania fields, $1.90 in Tephrosia fields, and $1.10 in non-fertilised maize fields. The results do not support the popular notion that agroforestry practices are more labour intensive. It suggests that in addition to total labour use requirements, the temporal distribution of labour use during the farming season is important. It is concluded that agroforestry practices are profitable, and could contribute to food security.