A structured German-language interview was developed for the assessment of painful and non-painful phantom and stump phenomena after amputation. The aim was a thorough assessment of the quality, quantity and time course of these phenomena, which is of scientific as well as therapeutic relevance. Each phenomenon was assessed using visual analogue scales as well as qualitative descriptors adapted from the McGill Pain Inventory and from literature reports. The factor structure and psychometric properties of the interview were evaluated in a sample of 139 upper and lower limb amputees. Test-retest coefficients were obtained in a subset of 20 amputees. As expected, all pain-related scales showed a two-dimensional internal structure with the factors "affective pain" and "sensory pain". For the non-painful phantom sensations, three factors "general/kinesthetic phantom sensations", "phantom movements" and "paresthesias" were obtained, while for non-painful stump sensations only one general factor emerged. The internal consistency was high with respect to the pain-related scales and was still satisfying for the scales that cover nonpainful phantom and stump phenomena. All scales have sufficient validity. Test-retest coefficients suggest a satisfactory stability of all scales that assess present phenomena, while the stability of the retrospective scales is markedly lower and in some cases insufficient. The phantom and stump phenomena interview is a highly reliable and valid instrument to assess present perceptual phenomena after amputation. Only the included retrospective scales apparently show low stability scores over time. This raises the more general question of the validity of retrospective pain reports.