This study is concerned with the effect of external turbulence on the decay of vortices. Single vortices and vortex pairs were generated with wing(s) mounted in the sidewalls of a wind tunnel. The distance between the two vortices could be adjusted such that they just touched each other or overlapped. The intensity of the turbulence could be controlled with a turbulence grid. The development of the vortex was measured at a number of downstream stations with particle image velocimetry for a range of wing settings. The results indicate that the single vortex can be described by the ‘two length scales’ model of Jacquin, Fabre & Geffroy (AIAA, vol. 1038, 2001, p. 1). A vortex core, which decays like a Lamb–Oseen vortex, is embedded in a region with an almost constant radius and a much lower azimuthal velocity that obeys a ~r−β power law, with r being the radius measured from the vortex centre. For the turbulence levels and the test section length used in this study, the single-vortex behaviour is independent of the external turbulence and in contrast with the decay of the vortex pair that strongly depends on the external turbulence. In the initial stages of the vortex pair evolution, the vortices decay due to cancellation of vorticity at the symmetry plane. At a later stage, Crow oscillations are observed, followed by a breakdown of the vortices. This vortex breakdown might be due to direct turbulent action. The observed behaviour is in agreement with the theoretical model of Crow & Bate (J. Aircraft, vol. 13, 1976, p. 476).