Using the Lyman break technique (e.g., Steidel et al. 1995), large samples of star-forming galaxies at $2 \lt z \lt 5$ have been identified and studied. These Lyman break galaxies (LBGs) are UV-luminous and thought to be similar to local starburst galaxies; they are relatively small (r$_{h}$=1-3 kpc), have relatively low mass (10$^{9.5-11}$M$_\odot$) and low extinction. However, their role in galaxy evolution is still debatable. Do all galaxies go through a Lyman Break phase? How biased is our view of galaxy evolution due to the Lyman break technique? Are LBGs the building blocks of larger systems or just small galaxies having their first burst of star-formation? Therefore, a census of the star-forming galaxy population as a function of time is needed in order to help us better understand how galaxies acquired their present morphology. In this contribution, I discuss the physical properties of a sample of UV-selected galaxies at intermediate redshifts. I conclude by showing that galaxies of all types, sizes and shapes are forming stars at intermediate-z. However, deep image and spectroscopy with large telescopes are needed in order to properly address their nature.