We investigated whether and how different imagined stress intensities modify responses in a stress coping inventory, the SVF78. In the original non-specific version of the test, subjects just decide for each item how probable the reaction presented corresponds to his or her way of reacting, when he/she is "...disturbed, irritated or upset by something or someone...". We compared this non-specific version of the SVF78 with three intensity variants containing reformulations of the introductory instruction: "When I am disturbed, irritated, or upset to a low degree/considerable degree/very high degree [ italics added] by something or someone...". Each subject filled in the non-specific version and two of the three intensity variants according to a balanced permutation design. Results showed good internal consistency as well as split-half-reliability for all subtests. The factorial structure was similar for the non-specific and the considerable- and high-intensity versions, but altered for the low-intensity version. The mean values changed monotonically with imagined stress intensity for some, but not all of the subtests of the SVF78, with the non-specific version matching best with the considerable-intensity level. It can be concluded that coping questionnaires like the one used in this study, based on a dispositional approach and asking for the normal way of reacting in stressful situations, are valid for a certain range of stress intensities mostly covering the considerable range of the intensity dimension.