Previous reports suggested a therapeutic response of lysosomal storage diseases such as Hurler syndrome following bone marrow transplantation. However, a clearer understanding of outcome has awaited long-term follow-up. We evaluated prospectively 11 consecutive patients with Hurler syndrome receiving marrow from an HLA-identical sib donor between September 1983-October 1988. Follow-up evaluations included assessment of donor engraftment by restriction fragment polymorphism analysis, determination of leukocyte alpha-L-iduronidase level, measurement of lumbar cerebrospinal fluid (CSF) pressure, computerized tomography (CT) of the brain, and psychometric testing. In this series there was a survival rate of 9/11 (82%) with all survivors showing complete (7 patients) or partial (2 patients) donor engraftment. Prospective longitudinal evaluation of the 9 surviving children, now 3.8-8.9 years posttransplantation (median 5.5) demonstrated persistence of previously deficient leukocyte alpha-L-iduronidase at levels reflecting the donor genotype and degree of donor engraftment. Urinary glycosaminoglycan excretion declined to near-normal within 5 months of donor engraftment. Prior to treatment, 7 of 8 children studied were found to have occult intracranial hypertension (lumbar CSF pressure > 20 cm CSF); however, all surviving children attained normal or near-normal pressure within 18 months of donor engraftment. Longterm follow-up CT imaging of the brain did not show progressive volume loss (cerebral atrophy) after donor engraftment. Of 9 survivors, 4 children having a developmental quotient (DQ, Mental Development Index on Bayley Scales of Infant Development) above 80 prior to transplantation subsequently maintained IQ scores above this level. However, 5 patients with lower pretransplant DQ scores now have significant cognitive deficits and attention deficit hyperactivity disorder. Progressive brain damage resulting from communicating hydrocephalus may be prevented by successful engraftment. Early transplantation of children with Hurler syndrome who have normal intelligence is likely to have the clearest benefit because long-term intellectual outcome will be limited by brain damage which has occurred prior to treatment.