In a prospective study of non-operative treatment of de Quervain tenosynovitis, ninety-nine wrists of ninety-five consecutively seen patients who had this diagnosis had an injection of one milliliter of a 1 per cent lidocaine solution and one milliliter of a suspension containing forty milligrams of methylprednisolone acetate. Twelve patients (twelve wrists) were lost to follow-up. Of the remaining eighty-seven wrists, fifty-four (62 per cent) had a satisfactory outcome at a mean of eighteen months (minimum follow-up, twelve months). The duration of symptoms before treatment did not affect the outcome. The result in thirty-three wrists (38 per cent) was considered unsatisfactory. Thirty of these wrists were subsequently treated with operative release of the first dorsal compartment, and twenty-two (73 per cent) of the thirty were found to have a separate compartment for the extensor pollicis brevis. The prevalence of a separate compartment is significantly higher than that in the general population, as shown in anatomical studies of cadavera.