This study investigates the effects of radon (plus CO2) baths on RA in contrast to artificial CO2 baths in RA rehabilitation using a double-blinded trial enrolling 134 randomised patients of an in-patient rehabilitative programme (further 73 consecutive non-randomised patients are not reported here). The outcomes were limitations in occupational context/daily living (main outcome), pain, medication and further quantities. These were measured before the start, after the end of treatment and quarterly in the year thereafter. Repeated-measures analysis of covariance (RM-ANCOVA) of the intent-to-treat population was performed with group main effects (GME) and group x course interactions (G x C) reported. Hierarchically ordered hypotheses ensured the adherence of the nominal significance level. The superiority of the radon treatment was found regarding the main outcome (RM-ANCOVA until 12 months: p(GME) = 0.15, p(G x C) = 0.033). Consumption of steroids (p(GME) = 0.064, p(G x C) = 0.025) and NSAIDs (p(GME) = 0.035, p(G x C) = 0.008) were significantly reduced. The results suggest beneficial long-term effects of radon baths as adjunct to a multimodal rehabilitative treatment of RA.