The aim of the study was to assess the role of illness perception in adaptation to chronic disease among patients with relapsing-remitting multiple sclerosis (RRMS). The differences between the obtained configurations of the illness perception components during four measurements and the model of predictions of the values of adaptation indicators, i.e. depression, anxiety and quality of life during subsequent measurements, were analyzed. Illness representation was assessed at baseline via the Illness Representation Questionnaire–Revised. The adaptation indicators–anxiety, depression (measured by HADS) and quality of life (measured by MSIS-29) were measured at baseline and three more times over a five-year period. The k-means cluster analysis (with two-way and repeated measures ANOVA) was conducted in a group of 90 patients (48.89% women and 51.11% men). Subsequently, the mean values of depression, anxiety, physical and psychological quality of life were compared between the clusters using the Kruskall-Wallis test. Finally, a cross-lagged panel modeled for HADS and MSIS-29 subscales in each measurement occasion (T1-T4). Three different illness perception clusters (Anxious, Realistic and Fatalistic Illness Perception named AIP, RIP and FIP) were composed which differentiated the depression, anxiety, quality of life level and age. FIP showed the lowest adaptation outcomes with small differences between AIP and RIP. It was also significantly characterized by the highest age. The positive adaptation indicators were related to the RIP cluster. The model presented rather satisfactory fit (χ2(48) = 81.05; CFI = .968; TLI = .925; SRMR = .050) with slightly inflated RMSEA = .087 (90%CI .053-.120). Based on initial measurements of individual characteristics, it was possible to predict the functioning of patients after several years. For patients with AIP, the covariance of anxiety and depression was significant, for patients with RIP–depression and anxiety, and for patients with FIP–depression. In addition, each of the variables was a predictor of subsequent measurements in particular time intervals, illustrating the dynamics of changes. Results highlight that illness perceptions formed at the beginning of RRMS are important for the process of adaptation to the disease. Moreover, they showed the differences between the adaptation outcomes supporting the idea that a cognitive representation might be important for the level of psychological functioning.