Measurements of London penetration depth are a sensitive tool to study multi-band superconductivity and it has provided several important insights to the behavior of Fe-based superconductors. We first briefly review the "experimentalist -friendly" self-consistent Eilenberger model that relates the measurable superfluid density and structure of superconducting gaps. Then we focus on the BaFe 2 As 2derived materials, for which the results are consistent with 1) two distinct superconducting gaps; 2) development of strong in-plane gap anisotropy with the departure from the optimal doping; 3) appearance of gap nodes along the c−direction in a highly overdoped regime; 4) significant pair-breaking, presumably due to charge doping; 5) fully gapped (exponential) intrinsic behavior at the optimal doping if scattering is removed (probed in the "self-doped" stoichiometric LiFeAs); 6) competition between magnetically ordered state and superconductivity, which do coexist in the underdoped compounds. Overall, it appears that while there are common trends in the behavior of Fe-based superconductors, the gap structure is non-universal and is sensitive to the doping level. It is plausible that the rich variety of possible gap structures within the general s ± framework is responsible for the observed behavior.