We demonstrate how a combination of different elements can jointly provide plausible long-term trends for calorie intakes, crop yields and land use in Computable General Equilibrium (CGE) analysis. Specifically, we depict household demand based on a MAIDADS demand system estimated based on cross-sectional data. In order to control for calorie intake we first regress calorie intake on per capita income and construct a Leontief inverse to derive implicit calorie intakes from the final consumption of processed food. This allows jointly shifting preferences of the MAIDADS system by updating commitment terms and marginal budget shares, to arrive at plausible per capita calorie intakes during baseline construction. We control yields based on exogenous projections which we also use to parameterize our land supply functions. The contribution of the different elements is evaluated by comprising key developments in baselines up to 2050 constructed with different model variants