The usefulness of Whatman seed-test and No. 17 papers has been examined for purification and recovery on a preparative scale of the toxin present in residues of low toxicity from the Schantz alumina chromatographic method for purification of clam poison. A method was developed to minimize introduction of impurities from the papers into fractions recovered after chromatography whent-butanol: acetic acid: water (2:1:1) was used as developer. Seed-test paper gave better separation of the toxin from other components of the residues than did No. 17 paper. Residues of bio-assay as low as 300 MU/mg were found to be capable of enrichment by the method. On a scale of 500 mg per 18 × 22-in. sheet of pretreated seed-test paper, using residues of toxicity 1400 MU/mg, a 67% recovery of toxin enriched fourfold to 5800 MU/mg was achieved in four passes. Increased loading of the paper led to a decrease in both enrichment per pass and over-all recovery of toxin. The paper chromatographic properties, toxicity, and infrared spectrum of the recovered poison agree with those reported by Schantz and co-workers for pure clam poison dihydrochloride, but the specific rotation is lower by 25%. Four possible alternative explanations for the observed anomalous rotation are discussed.