The article reports on a study in which elements of a group-oriented reconditioning programme for patients with low back pain were incorporated into the routine of an in-patient orthopaedic rehabilitation clinic. The specific elements of the new programme consisted of stable group structures during the whole stay of 3 to 4 weeks, and of 3 to 7 hrs. walks in hilly ground three times a week. The effects of this programme were tested against a standard programme with a mix of passive and active elements using a controlled study design. 92 persons participated in an experimental group and 81 persons in a control group with no significant differences found between the two groups at admission. Effects were measured by means of a physicians' questionnaire and a multidimensional patients' questionnaire (IRES) answered at admission, at discharge, and at six- and twelve-month follow-up. The results of a two-factorial analysis of variance with repeated measures showed that the interaction between group and time on the summary score of the IRES was not significant, although the experimental group showed somewhat better effects at all times of measurement. The discussion focuses on the reasons for this result, among which the unexpectedly good effects in the control group are named, as well as certain difficulties with the implementation of a strongly activity-oriented programme into the course of a "normal" rehabilitation clinic.