The decrease of saving rate in Colombia has become not only a long run trend but also has exhibited an important structural break in the nineties. Appealing to an alternative hypothesis, that of thelabor market, in this work it is built a model including unemployment rate andgross rate of participation, besides per capita consumption and labor income. General Method of Moments estimations show that the effects of labor income and consumption are statistically the same, which not only strongly suggest the fulfilling of permanent income hypothesis, but also leave a high explicative power tolabor market variables, where unemployment rate has the highest significative effect and, furthermore, the structural break present in this last variable during nineties seems to be well encompassedwith the structural break present in saving rate. Thus, the empirical evidence embodied in the estimations suggests that an improvement in labor market conditions leading to unemployment reductions willhelp significantly recovering the saving rate in Colombia.