Fifty years after its widespread recognition, a significant minority of patients with carpal tunnel syndrome continue to experience poor outcomes from treatment. Much current treatment is supported by an inadequate or nonexistent evidence base. Surgical decompression, often considered the definitive solution, gives excellent results in only 75% of cases in ordinary practice and leaves 8% of patients worse than previously. The only other interventions that are clearly of benefit are neutral-angle wrist splinting, with a success rate of 37%, and steroids, which are better given by local injection than as oral treatment. The initial response rate to injection is 70% but there are frequent relapses. Nevertheless, these conservative treatments have a negligible incidence of serious complications and should be used more widely until surgical failures can be reduced to similar levels.