OBJECTIVE In mid 1989, leaded petrol was replaced by unleaded petrol to reduce lead toxicity in petrol sniffers in Maningrida, a remote Aboriginal community in Northern Australia. RETROSPECTIVE REVIEWS: Hospital admissions between 1987 and 1992 due to petrol sniffing were compared for Maningrida and a community using only leaded petrol. RESULTS Admissions from Maningrida decreased significantly following introduction of unleaded petrol (chi 2 on 2df = 22.25, p < 0.001). Lead and hydrocarbon exposures were also compared for three groups from Maningrida (27 sniffers using only unleaded petrol; 15 exsniffers and 13 nonsniffers) and 24 individuals admitted to hospital for petrol sniffing related illness from other communities using only leaded petrol. Median blood lead levels for hospitalized sniffers (using only leaded petrol). Maningrida sniffers (using only unleaded petrol), exsniffers and nonsniffers were 5.06, 1.87, 1.24 and 0.17 microM/L respectively. There were significant differences between blood lead level, delta-aminolevulinic acid dehydratase activity, and free erythrocyte protoporphyrin for sniffers of leaded and unleaded petrol, whereas these indices were not significantly different for current sniffers and exsniffers in Maningrida. Hydrocarbons were only detectable in the blood of active sniffers (toluene < or = 0.5 micrograms/mL; benzene < or = 0.17 micrograms/mL blood; n-hexane not detected). CONCLUSIONS The elimination of tetraethyl lead from petrol resulted in a significant decrease in hospitalization of petrol sniffers.