The interlayer diffusoin barrier for Fe/Fe(100) deposition is estimated by comparing the results of kinetic Monte Carlo simulations with experimental results in the first few monolayers of growth. We find that, in contrast to previous theoretical estimates for other systems, the step barrier for Fe/Fe(100) is small in comparison with the activation energy for diffusion on a flat terrace (0.454 eV). Our results resolve a long-standing controversy and provide quantiative support for the conjecture that the existence of mounds in this system is due to a finite positive step barrier.