It is shown that attempts to accurately deduce the magnetic penetration depth λ of overdoped BaFe1.82Co0.18As2 single crystals by transverse-field muon spin rotation (TF-µSR) are thwarted by field-induced magnetic order and strong vortex-lattice disorder. We explain how substantial deviations from the magnetic field distribution of a nearly perfect vortex lattice by one or both of these factors is also significant for other iron-based superconductors, and this introduces considerable uncertainty in the values of λ obtained by 74.25.Ha, 76.75.+i TF-µSR is routinely used to determine the magnetic penetration depth λ of type-II superconductors in the vortex state for the purpose of obtaining indirect information on the energy gap structure [1]. The magnetic field distribution n(B) in the sample is determined by detecting the decay positrons from implanted positive muons that locally probe the internal fields, and λ is subsequently determined by modeling the contribution of the vortex lattice (VL) to n(B). However, even in conventional superconductors the VL contribution is not known a priori, and one must rely on phenomenological models to deduce what is really an "effective" penetration depthλ. One reason for this is that only cumbersome microscopic theories account for the effects of low-energy excitations on n(B) [2]. Extrapolating low-temperature measurements ofλ to zero field to eliminate the effects of intervortex transfer of quasiparticles, as well as nonlocal and nonlinear effects, has been demonstrated to be an accurate way of determining the "true" magnetic penetration depth λ [3,4]. Yet an underlying assumption is always that the VL is highly ordered and that other contributions to n(B) are relatively minor. The purpose of this Letter is to point out that this is not the case in many of the recently discovered iron-based superconductors, making a reliable determination of λ by TF-µSR extremely difficult.Here we report on representative TF-µSR measurements of BaFe 1.82 Co 0.18 As 2 (T c = 21 K) single crystals grown from a FeAs flux, as described elsewhere [5]. High-statistics TF-µSR spectra of 20 million muon decay events were collected in magnetic fields H = 0.02 T to 0.5 T applied transverse to the initial muon spin polarization P (t = 0), and parallel to the c-axis of the crystals. The TF-µSR signal is the time evolution of the muon spin polarization, and is related to n(B) as followswhere γ µ is the muon gyromagnetic ratio. Generally, the TF-µSR signal is fit in the time domain, with the inverse Fourier transform or "TF-µSR line shape" given byproviding a visual approximation of the internal field distribution. The field distribution of a perfectly ordered VL is characterized by sharp cutoffs at the minimum and maximum values of B(r), and a sharp peak at the saddlepoint value of B(r) [1]. These features are not observed in polycrystalline samples, where the orientation of the crystal lattice varies with respect to H, but are observed in single crystals when a highly-ordered VL exists and other co...