Little work has been yet done to analyse if e-learning is an efficiency way in economic terms to produce higher education, especially because there are not available data in official statistics. Despite of these important constrains, this paper aims to contribute to the study of economic efficiency of e-learning through the analysis of a sample of e-learning universities during a period of time (1997-2002). We have wanted to obtain some empirical evidence to understand if e-learning is a feasible model of providing education for universities and which are the variables that allow for feasibility attainment. The main findings are: 1) that the rise of the number of students enrolled is consistent with increasing labour productivity rates; 2) that cost labour savings are explained by the improvement of universities¿ economic efficiency (or total factor productivity); and 3) that improvement of total factor productivity in e-learning production is due to the attainment of scale economies, but also to two organisational innovations: outsourcing processes that leads to the increase of variable costs consistent with decreasing marginal costs, and the sharing of assets¿ control and use that allow for a rise in assets rotation.