Costs and benefits associated with erosion reduction and adoption of soil conservation practices for a representative farm in a watershed in Honduras are examined in a linear programming framework. Special attention is paid to income-soil loss tradeoffs, income-risk tradeoffs, and on the effect of different farmers' planning horizons on net farm income. A representative farm model for the area was constructed to achieve the objectives of the study. Twelve farmers in the region were surveyed, crop budgets were prepared, and soil loss values were calculated to provide the information required to construct this representative farm. A linear programming model that maximizes net farm income is used to examine the effect of different soil loss levels on farm income. A MOTAD model that minimizes deviation in income (risk) is used to determine risk levels while income and/or soil loss levels restrictions are imposed. Results indicate that considerable reductions in the amount of soil loss can be achieved in the study area. Erosion is reduced from 328.24 ton./mn./year to 6.56 ton./mn./year1 when constraints are imposed on the model. The reduced erosion lowers income from L.5929.24/year for high erosion rates to L.2825.8l/year for low erosion rates. Low levels of soil erosion are achieved at the expense of higher levels of risk. High levels of income are associated with high levels of risk regardless of whether soil loss constraints exist or not. Small differences in income exist among the four planning horizons analyzed. The best soil conservation practices for this region turned out to be the cultivation of coffee on the highest slopes, the use of live barriers and terraces, and the use of conventional and minimum tillage. / Master of Science