This thesis seeks to rebuild and extend the primal approach in production analysis by: a) developing a new paradigm for agricultural production analysis that acknowledges the relevance of the agronomic aspects of the production processes; b) defining factor demand in agricultural production from a primal, technical perspective. The theory and approach developed are used to address aspects of environmental and economic sustainability of agriculture in the Netherlands. This research advocates that production models should respect and incorporate prior knowledge into production analysis. The research recognizes that inputs used in agricultural production have distinct functionalities and this must be accounted for explicitly in the model. Input functionality is first defined as productive inputs, damage-abating inputs, and inputs that have both functionalities. Then the research further generalizes this distinction by defining a dichotomy of growth inputs and facilitating inputs. Growth inputs are directly involved in the biological process of crop growth and include land, seed, fertilizer, and water; facilitating inputs help create or alter growth conditions, under which growth inputs take effect, and include labor, capital, and pesticides. Based on this dichotomy, a conceptual framework is developed. The framework connects agronomy and economics, presenting a new paradigm for production analysis. The empirical analyses use data from Dutch arable farms. Test results reject the traditional translog model, in favor of the new model proposed in this research. Next, the research addresses input demand from a primal, technical perspective. A capital requirement model is proposed, and the factors that affect the capital requirement in production are investigated. Using the capital requirement model, this research defines a concept of excess capital and finds excess capital widely exists in Dutch arable farming. The presence of excess capital implies mismeasurement and simultaneity of capital in production. Its implications in econometric analysis of production are demonstrated with a production frontier model. Results show that models not addressing these problems would yield biased and inconsistent efficiency estimates. In addition to the separate studies of product supply and input demand, this research further combines the two aspects and conducts an analysis on the total factor productivity growth in Dutch agriculture, focusing on the impacts of capital structure, investment, and governmental subsidies