On the light of the history between the years 1954-2002 (for relations that were found in this period, the estimations are actualized with information to 2003), period that includes an increase in income growth volatility for Colombia, the objective of this work is, in the context of different consumption theories, to find empirically what macroeconomic variables have a long run relationship with household consumption and quantify its incidence.Additionally and according with the estimated relations, evaluate its goodness of fitness and the contribution of this to the theoretical discussion.The results confirm for the Colombian case the low relevance of demographic structure, the negative effect of interest rate and the relevant roll of wealth, disposable income and financial system.On the long term households with a higher access to the financial system tends to save.