An alternative methodology is described for evaluating the effect of different vegetation covers on streamflow regulation in small watersheds. Toward this aim, statistical procedures are employed, including hypothesis testing on regression slopes, which has not been used before for this purpose, and it is based upon easily obtained hydrological information. Subsequently, its applicability is demonstrated for La Murcielago stream watershed (Piedras Blancas, Antioquia) which is covered by Pinus patula and Cupressus lusitanica forest plantations, and natural forest. Results indicate significant differences between pine plantation, and cypress and natural forest covers, the latter with the smaller slope and therefore whit a higher regulation capacity on water yield.