We compared the extent of soil erosion between maize and agroforestry farming systems through monitoring runoff and sediment loss processes on six representative plots along a mountain elevation gradient. Two plots, one under maize and another under agroforestry farming system, were established in each elevation zone. Under agroforestry, study plots with similar canopy types and composition were selected. We installed one erosion trap in each study plot. Erosion traps measuring 10 m length and 3 m wide were installed in each study plot. The erosion plot was carefully delineated using joined cement blocks (30 cm high.) without disturbing the soil surface and the vegetation. At the lower part of the trap, a triangle shape covered by a plastic sheet channeled the runoff through a PVC pipe to a 150 liter plastic barrel (Fig. 2). Rainfall gauges ware installed in the vicinity of each plot (50m) outside of tree and banana canopies. Since transport logistics could not allow us to attend erosion plots after every rainfall event, we scheduled our data collection exercise after every two days. Rainfall volume was measured by using a measuring cylinder whereas the volume of runoff was computed based on its height inside the barrel and the cross-sectional area of runoff and the bottom of the barrel. The runoff mixture was vigorously stirred before collection of 0.5 L sub-samples. Barrels were then cleaned up for the recording of the following runoff events. The collected samples were oven dried at 110 °C and reweighted until constant weight. Runoff and rainfall data were recorded on a cumulative basis for the whole observation period.