BACKGROUND: Mycosis fungoides (MF) is the most common skin lymphoma. The aetiology of MF remains unknown, and no therapy has to date significantly altered patient survival. OBJECTIVE: The present study examines trends in survival of MF patients in a well-defined population-based disease group, namely patients registered over a 40-year period at the Thames Cancer Registry, Southeast England. METHODS: The Thames Cancer Registry is a population-based registry, covering a population of approximately 14 million people. Data were taken from the Surveillance, Epidemiology and End Results cancer registry programme and the National Centre for Health Statistics. The database was used to identify cases of MF diagnosed between 1961 and 2000. A total of 985 records were identified, 821 (83%) of which had complete information on age, sex, year of diagnosis and area of residence. The observed and relative survivals of patients diagnosed during the periods 1971-1975, 1981-1985 and 1991-1995 were examined over a 5-year period of follow-up, using the relsurv Stata program to perform Cox proportional hazard analysis. RESULTS: A total of 821 MF eligible patients were available with a median follow-up of 4.3 years and a maximum follow-up of 30 years. The overall 5-year relative survival rate was 80%, and there was marked improvement between 1971 and 1981. The prognostic factors leading to a significantly poorer survival were high age, male sex, the presence of the Sézary syndrome, the use of hormone treatment and radiotherapy. CONCLUSIONS: A statistically significantly better survival over the last 20 years was found. The prognosis is generally good for most patients but not all. The best survival was seen for the female patients under 45 years of age without the presence of the Sézary syndrome. This difference in survival may be partly due to a difference in the disease stage or different treatment, or to both.