According to our present understanding of the process of structure formation, halos (gravitationally bound structures) of different masses trace the underlying mass density in different ways. This is know as halo bias. Since the properties of galaxies are also different depending on the halos masses, merger histories, tidal fields, etc., they also constitute different tracers of the density field. I will explore how we can exploit multi-tracer maps of large-scale structure, from the optimal estimation of power spectra to enhanced constraints on certain parameters of cosmological interest, such as the nature of dark energy and modifications of General Relativity.