Hospital-based proton facilities may represent a major advance in radiation therapy, in part because of excellent dose distributions around the tumor, and in part because of the potentially lower whole-body dose compared with photon radiotherapy. Most current proton beams are spread out to cover the tumor using beam scattering and collimation techniques (passive scattering); this will necessarily result in an extra whole-body neutron dose, due to interactions of the protons with the scattering and collimating beam elements. However, the clinical significance of this whole-body low-dose neutron exposure has remained controversial. The number of proton facilities worldwide is increasing rapidly, and most of these facilities are/will be based on passive scattering. Thus it is important to assess and, ideally, minimize, the potential for second cancer induction by secondary neutrons. We discuss here the neutron doses involved, and the associated potential second cancer risks, with an emphasis on the uncertainties involved.