We report the findings in 60 patients with secondary progressive multiple sclerosis who had monthly brain MRI studies for 4 months (one baseline and three follow-up scans). The purpose was to define the short-term MRI natural history in a large cohort with secondary progressive disease and to ascertain its relationship with other clinical and MRI features. The patients were participating in either a natural history study or the placebo arm or non-treatment phase of a therapeutic trial. The cohort had clinical features typical of secondary progressive disease: thus, all had moderate or severe locomotor disabilities [Expanded Disability Status Scale (EDSS), score 3.5-8], with a median disease duration of 12 years. There was equal representation of males and females. During the 3 months of follow-up there was a total of 362 new enhancing lesions seen in 42 patients, and there were 24 relapses in 20 patients. There was no correlation between new enhancing lesions and age at study entry, age of disease onset, gender disease duration or EDSS, but there was a strong correlation with the number of enhancing lesions on the baseline scan (r = 0.65, P < 0.0001) and subsequent activity. There was a non-significant trend for higher numbers of new enhancing lesions in those having relapses during the 3 months of scanning (P = 0.14) or in the preceding 6 months (P = 0.06). The 34 patients who did not relapse in either period had significantly fewer new active lesions (P = 0.02) than those who relapsed at some stage during the 9 months. Nevertheless, considerable activity was seen in the non-relapsing cohorts: there was a mean of 3.5 (median 2) new enhancing lesions in those not relapsing during the 3 month study, and 5.5 (median 2) in those not relapsing in the previous 6 months. We conclude that short-term MRI activity is generally high in secondary progressive disease, confirming a useful role for the technique in exploratory trials. Further work should concentrate on elucidating the mechanisms of secondary progression by longer term follow-up studies of larger cohorts using multiple MRI and clinical measurements.