Errors in atmospheric transport of tracers lead to errors in estimates of tracer fluxes based upon concentration observations. Typically, such inverse methods either neglect transport errors or only assess their effects roughly. We describe a method to quantitatively account for transport errors by incorporating uncertainties in winds into stochastic motions of air parcels. The magnitude of errors in wind fields, as well as their spatiotemporal covariances, are determined by direct comparison of assimilated winds to radiosonde observations. These statistics of transport errors are propagated through stochastic motions of air parcels in a Lagrangian model (STILT). We illustrate this method by conducting an inverse analysis using simulated CO2 observations over the continent and examine the effect of transport errors on estimates of regional terrestrial carbon fluxes. The inverse analysis demonstrates that transport errors can cause significantly biased estimates. We show that the proposed method properly accounts for these errors.