Surface enhanced Raman spectroscopy (SERS) of NO on cold-deposited Cu yields only bands of the dissociation products of NO, but not the stretching bands of NO. In contrast, infrared reflection-absorption spectroscopy (IRRAS) on similarly prepared samples displays strong NO stretching bands and only a very weak band of N 2 O from multilayer NO. The differences can be explained neither by thermal or photodesorption nor by photochemical effects in the laser focus. Whereas the SERS results may be explained by the action of special sites, it is still unclear why the IRRAS signal from the dissociation products of NO at the 'catalytically active sites' is below the noise level.