Black Carbon (BC) aerosols substantially affect the global climate. However, accurate simulation of BC atmospheric transport remains elusive, due to shortcomings in modeling and a shortage of constraining measurements. Recently, several studies have compared simulations with observed vertical concentration profiles, and diagnosed a global-mean BC atmospheric residence time of <5 days. These studies have, however, been focused on limited geographical regions, and used temporally and spatially coarse model information. Here we expand on previous results by comparing a wide range of recent aircraft measurements from multiple regions, including the Arctic and the Atlantic and Pacific oceans, to simulated distributions obtained at varying spatial and temporal resolution. By perturbing BC removal processes and using current best-estimate emissions, we confirm a constraint on the globalmean BC lifetime of <5.5 days, shorter than in many current global models, over a broader geographical range than has so far been possible. Sampling resolution influences the results, although generally without introducing major bias. However, we uncover large regional differences in the diagnosed lifetime, in particular in the Arctic. We also find that only a weak constraint can be placed in the African outflow region over the South Atlantic, indicating inaccurate emission sources or model representation of transport and microphysical processes. While our results confirm that BC lifetime is shorter than predicted by most recent climate models, they also cast doubt on the usability of the concept of a “global-mean BC lifetime” for climate impact studies, or as an indicator of model skill.