This paper deals with human capital and firm-level performance in agriculture. In this sector firms are comparatively small. Management, capital provision and labor supply are concentrated on one individual and his immediate family. Many tasks have to be done, and decisions have to be worked out and carried through. Mistakes are made, not only on what crop variety to plant, but also on how much land and capital to allocate to different activities. Difficulty in decision-making is compounded by the impact of climatic variability on resource productivity. Indeed, in agriculture, a given input bundle can result in very different output levels according to the level at which random factors operate. Uncertain resource productivity coupled with imperfect capital markets result in a "bad" year seriously compromising the flow of funds to the household. 

Attention is focused here on the "pradera pampeana" (prairie of the Argentine pampas) , which comprises the main production area of Argentina. To the author's knowledge, only one previous paper (Gallacher, 1999) has addressed the issue of human capital-agricultural production in Argentina. This paper advances on previous work in one important aspect: the data used constitutes a 10 % random sample drawn from records of the 1988 Agricultural Census (Censo Nacional Agropecuario 1988). Some 5300 firms constitute the data set. The questions to be addressed are: (1) Does farmer education affect production efficiency (2) Do returns to education depend on firm technology, and lastly, (3) What are the economic returns associated to different educational levels, and how do these returns compare to those found, in Argentina, in the non-agricultural sector. The paper also attempts to make inferences regarding the interaction between human capital and returns to scale to (conventionally measured) factor inputs.