The 4-level questionnaire to assess leisure-time physical activity (PA) originally described by Saltin and Grimby has been extended to a 6 (or 7)-level questionnaire, especially for the elderly or other persons with a low level of PA, expanded to include household activities. There are currently two main versions, one used in Swedish, Danish, and Norwegian studies, and one with further modifications used in Finnish studies. In this review, predictive validity for mortality and different aspects of physical performance are demonstrated. In a study regarding concurrent validity in one of the Finnish versions, significant correlations with accelerometer-based PA variables are shown. The scale has been used in a number of studies concerning effects of different types of intervention, such as physical training and increased PA, or to describe the PA level in the studied cohort. In presenting and analyzing the data, different combinations of PA levels have been used. As there are major differences between the versions, validity studies cannot easily be transformed between them. Thus, it is suggested that a common consensus be reached with respect to details of the questionnaire.