The intervention, “Rheumates @ Work,” was an Internet-based program lasting 17 weeks, combined with 4 group sessions. It contained the following elements: 1) health education: teaching the patient the fundamentals of cause and effect of JIA in order to improve disease management, 2) benefits of PA are explained and barriers toward PA are identified, 3) self-efficacy and perceived affect of PA are identified and reinforced, 4) influence of family and school is recognized and used to promote PA, 5) PA options in daily life are explored and encouraged, and 6) smart goals are set: patients have to better tune and define specific, measurable, attainable, realistic, and timely goals, such as “I am going to cycle to school 3 times a week instead of going by car, for the coming 2 months.” All of these elements are based on the health promotion model by Pender, which identifies moldable determinants of PA behavior (17). Films, animations, spoken texts, puzzles, brain twisters, and assignments were used and presented in dosed quantities. The program started with an interactive group session for patients and their parents, where an individual goal was set by the patient depending on their joint status and PA level. A contract expressing the will to participate was signed. Then the weekly Internet sessions started, and every week the patients worked on a specific theme: week 1 = what is JIA; week 2 = how to tackle problems; week 3 = energy and fitness; week 4 = benefits and barriers of PA; week 5 = pain and PA; week 6 = setting goals; week 7 = how to train; week 8 = second group session about barriers and benefits of PA; week 9 = taking responsibility; week 10 = chat box; week 11 = third group session about how to ask help from parents, friends, and school; week 12 = support and being physically active together; week 13 = how to be open over your JIA and ask for help; week 14 = how to deal with setbacks; week 15 = rewarding yourself; week 16 = retrospective; and week 17 = final group session: dealing with setbacks and adjusting goals. The patient received a reminder by e-mail when he/she did not complete the assignment. A feedback loop was built into the program to verify whether the patient had read the theory and made the assignment. During the entire week, the patient could ask questions by mail. After registration online, each patient had their own personal Web page that provided information about their current state of PA, fitness, and joint status. This information was measured at baseline. On this page, a small figure was depicted visualizing all of the joints by having 3 possible colors: green, orange, or red, depending on the physical examination of the rheumatologist. Green indicated that the joint could and should be normally used with respect to normal physical limitations. Orange indicated a joint with some degree of active arthritis, whereby light activities of daily life were encouraged but strenuous, and high-intensity activities were discouraged. Red indicated a joint with joint damage and therefore restricted function. During the intervention, the patient received the normal standard treatment for the JIA, which means that most patients were only seen at the beginning and at the end of the trial by a pediatric rheumatologist (WA). For those patients followed up in between, changed joint status was adapted at the personal Web page. The patients were free to perform their sport and leisure activities. Other physical training programs were allowed. Cognitive behavioral training focused on PA outside of the study was not permitted.