Anti-inflammatory drugs are effective on relapses, but neuroprotective agents to prevent disability are still unavailable. Uric acid has neuroprotective effects in experimental models including encephalomyelitis and appears to be involved in multiple sclerosis. Oral administration of inosine, a precursor of uric acid, increases serum uric acid levels and is well tolerated. Our objective was to test the possibility that a combination therapy associating an anti-inflammatory drug (interferon beta) and an endogenous neuroprotective molecule (uric acid) would be more effective than interferon beta alone on the accumulation of disability. Patients with relapsing-remitting multiple sclerosis on interferon beta for at least 6 months were randomized to interferon beta + inosine or interferon beta + placebo for 2 years. The dose of inosine was adjusted to maintain serum uric acid levels in the range of asymptomatic hyperuricaemia (<or=10 mg/dl). The primary end points were percentage of patients with progression of disability and time to sustained progression (Kaplan-Meier analysis). The combination of interferon beta and inosine was safe and well tolerated but did not provide any additional benefit on accumulation of disability compared with interferon beta alone. We conclude that endogenous neuroprotective mechanisms recently identified in multiple sclerosis are complex and uric acid does not reflect the entire story.