OBJECTIVES: To describe variations in fatigue over the course of 2 years in a sample of persons with multiple sclerosis (MS), and to investigate the predictive value of the following variables on variations in fatigue: sex, age, sense of coherence, living with a partner, living with children, work status, immunomodulatory treatment, mood, disease severity, disease course, time since diagnosis and time. METHODS: Every 6 months, 219 outpatients at an MS specialist clinic were assessed using the Fatigue Severity Scale (FSS). Predictive values were explored with Generalised Estimating Equation employing proportional odds models; FSS scores were categorised as non-fatigue, borderline fatigue or fatigue. RESULTS: FSS scores varied significantly (p = 0.02); 54% changed FSS category one or several times, 27% were persistently fatigued and 19% persistently non-fatigued. Independent predictors of increased fatigue were depressive symptoms, weak/moderate sense of coherence, living with a partner and not working. Furthermore, moderate disease severity predicted increase when combined with >10 years since diagnosis or a progressive course. Independent predictors of decreased fatigue were no depressive symptoms, strong sense of coherence, living alone and working. Moreover, mild and severe disease predicted a decrease when combined with >10 years since diagnosis, and mild severity combined with a progressive course. CONCLUSION: Mood, sense of coherence and living with a partner were independent predictors of fatigue in persons with MS. In addition to monitoring disease related variables, health related services should apply a broad range of approaches and repeatedly assess fatigue in persons with MS, to provide preventive care and appropriate interventions.