The properties of planets orbiting a component of a binary or multiple-star system appear different from the ones of planets orbiting isolated stars (Udry et al. 2004). The effect of binarity has, however, still to be better characterized and quantified. In this contribution, we describe a VLT/NACO adaptive optics survey aimed at pointing out the potential effect of stellar duplicity on planet formation and evolution, by comparing the faint-companion statistics of similarly selected subsamples of stars with and without planets. The statistical results of the survey will allow us to possiblydiscriminate between the proposed models of planetary formation (accretion or disk instability), some of them predicting different influences of stellar duplicity. Preliminary results are already very interesting: 20 ± 7% of the “single” stars present companions within ~10″, whereas only 3 ± 3% equivalent companions are found close to stars hosting planets.