According to our present understanding, long gamma-ray bursts (GRBs) originate from the collapse of massive stars, while short bursts are caused by to the coalescence of compact stellar objects. Because the afterglow evolution is determined by the circumburst density profile, n(r), traversed by the fireball, it can be used to distinguish between a constant density medium, n(r) = const., and a free stellar wind, n(r) ∝ r-2. Our goal is to derive the most probable circumburst density profile for a large number of Swift-detected bursts using well-sampled afterglow light curves in the optical and X-ray bands. We combined all publicly available optical and Swift/X-ray afterglow data from June 2005 to September 2009 to find the best-sampled late-time afterglow light curves. After applying several selection criteria, our final sample consists of 27 bursts, including one short burst. The afterglow evolution was then studied within the framework of the fireball model. We find that the majority (18) of the 27 afterglow light curves are compatible with a constant density medium (ISM case). Only 6 of the 27 afterglows show evidence of a wind profile at late times. In particular, we set upper limits on the wind termination-shock radius, RT, for GRB fireballs that are propagating into an ISM profile and lower limits on RT for those that were found to propagate through a wind medium. Observational evidence for ISM profiles dominates in GRB afterglow studies, implying that most GRB progenitors might have relatively small wind termination-shock radii. A smaller group of progenitors, however, seems to be characterised by significantly more extended wind regions. Appendices are only available in electronic form at http://www.aanda.org