The aim of this thesis was to study different aspects of the soft tissues of the shoulder joint. The variation in the quality of the tendons and ligaments can be explained by genetic factors. To test the hypothesis that collagen 1 α1 Sp1 polymorphism is related to the occurrence of cruciate ligament ruptures and shoulder dislocations, a total of 358 patients (233 patients with cruciate ligament ruptures and 126 with shoulder dislocations) were included in the study. We found a decreased risk of these injuries associated with collagen type 1 α1 Sp1 polymorphism. To study the mechanical properties of a better type of fixation of soft tissue to bone, 10 skeletally mature New Zealand white rabbits were operated bilaterally on the knees. The medial collateral ligaments were fixed by two types of plates one with a flat undersurface and the other with a pegged undersurface. After 4 weeks the force at failure, stiffness and energy uptake was almost double in the knees operated with the pegged plates. The prevalence and dysfunction of rotator cuff tears was investigated in 106 subjects who had never sought for their shoulder complaints, using Constant score, ultrasound and plain x-ray. The prevalence of full-thickness cuff tears was 30% (21% of all shoulders). The Constant score was lower in subjects with full-thickness tears. Partial-thickness tears and acromioclavicular joint osteoarthritis had no impact on shoulder complaints or Constant score. The subacromial index was lower for shoulders with full-thickness tears. Forty-eight patients with median age 56 years underwent subacromial decompression with or without acromioclavicular joint resection, investigated with MRI pre- and 3 months postoperatively. The Constant score and subjective shoulder value were measured preoperatively and at 3 and 6 months after surgery and even 2 years for subjective shoulder value. Two raters investigated the MRI. The results showed poor inter-rater reliability for MRI. However, both Constant score and subjective shoulder value improved over time. MRI is not a reliable method to study the capsular reaction after subacromial decompression due to high subjectivity of the radiologists.