Methyl n-butyl ketone (MnBK) and n-hexane produce a polyneuropathy in experimental animals. Metabolic studies have revealed that both of these compounds are metabolized to similar metabolites; that is 2-hexanol, 5-hydroxy-2-hexanone, 2,5-hexanediol, and 2,5-hexanedione. The latter two metabolites have been shown to be neurotoxins and the former two may be neurotoxic by virtue of the fact that both are metabolized to 2,5-hexanedione. The relative neurotoxicity of MnBK, n-hexane, and their metabolites was compared by administering equimolar doses of each compound by gavage to groups of male rats, 5 days/week over a 90-day period. A control group was given distilled water. The endpoint used to determine relative neurotoxicity was the time (days) required to produce clinical evidence of severe hindlimb weakness or paralysis. All test compounds produced both clinical and histologic evidence of neuropathy. Morphologically, this nerve damage was identical to that previously described following MnBK, n-hexane, and 2,5-hexanedione administration. The relative neurotoxicity of the test compounds in decreasing order of potency was: 2,5-hexanedione, 5-hydroxy-2-hexanone, 2,5-hexanediol, methyl-n-butyl ketone, 2-hexanol, and n-hexane. The neurotoxic potency was directly related to the amount of 2,5-hexanedione produced by each compound as determined by measuring the area under the serum concentration-time curve of 2,5-hexanedione. Additionally, atrophy of testicular germinal epithelium similar to that previously reported for 2,5-hexanedione occurred in the animals receiving 2,5-hexanedione, 2,5-hexanediol, 5-hydroxy-2-hexanone, MnBK, and 2-hexanol. n-Hexane administration had a lesser effect on the germinal epithelium than did the other compounds. Effects on body weight response and feed consumption paralleled the neurotoxic potency of each compound.