The known neuroprotective actions of melatonin could be due to its antioxidant or radical scavenging activity, or they could be due to specific interactions of the indole with its receptors. A study of structure–activity relationships may provide useful information when a validated macromolecular target has not been (or is not) identified. A set of indole derivatives, with changes in the 5-methoxy and acylamino groups, the side chain position and the lipophilic/hydrophilic balance, were selected and tested for their in vitro antioxidant potency in the ABTS (2,2¢-azinobis (3-ethylbenzothiazoline-6-sulfonic acid disodium salt) and thiobarbituric acid reactive substances (TBARS) assays and for their cytoprotective activity against kainate excitotoxicity on cerebellar cell cultures. No quantitative model was able to relate the potencies obtained in the two antioxidant assays, probably because they are related to different physico-chemical properties. However, the lipophilicity of the compounds and the antioxidant potency in the TBARS assay were linearly correlated. This may be due to improved access to the lipidic substrate, where the antioxidant action occurs. In the cytoprotection assay, most compounds showed potencies comparable with or lower than melatonin. An exception was N-[2-(5-methoxy-1H-indol- 2-yl)ethyl]acetamide (12), yielding, at 50 lm, percentages of cell vitality higher than 75%, while melatonin EC50 was 333 lm. No correlation was observed between cytoprotective and antioxidant potencies, nor with MT1 or MT2 receptor affinity. Compound 12 is a low-affinity antagonist at melatonin membrane receptors, and one of the most potent compounds in the antioxidant assays; its cytoprotective potency and the absence of agonist activity at melatonin membrane receptors make it a valid candidate for further investigations.