Abstract A recently developed technique was used to compare the local neurotoxicity of three white spirit (crude oil distillates) formulations. They differed in aliphatic and aromatic hydrocarbon concentrations. Formulations I and II contained over 10% n-nonane, whereas its concentration was negligible in formulation III with 17% aromatics. Formulation I contained no aromatics. Formulation I and II caused local axonopathy by neurophysiological and morphological criteria after 6 weeks of daily 3-h exposures for 5 days a week. No neurophysiological changes were caused by formulation III while morphological analysis revealed infrequent demyelinative foci after 6 weeks. Rats exposed to formulation II with 1% trimethylbenzene isomers showed urinary excretion of dimethylbenzoic acid isomers not detected in other rats thus providing a method for the biological monitoring of exposure.