The human malaria parasite Plasmodium falciparum is able to synthesize de novo pyridoxal 5-phosphate (PLP), a crucial cofactor, during erythrocytic schizogony. However, the parasite possesses additionally a pyridoxine/pyridoxal kinase (PdxK) to activate B6 vitamers salvaged from the host. We describe a strategy whereby synthetic pyridoxyl-amino acid adducts are channelled into the parasite. Trapped upon phosphorylation by the plasmodial PdxK, these compounds block PLP-dependent enzymes and thus impair the growth of P. falciparum. The novel compound PT3, a cyclic pyridoxyl-tryptophan methyl ester, inhibited the proliferation of Plasmodium very efficiently (IC(50)-value of 14 microM) without harming human cells. The non-cyclic pyridoxyl-tryptophan methyl ester PT5 and the pyridoxyl-histidine methyl ester PHME were at least one order of magnitude less effective or completely ineffective in the case of the latter. Modeling in silico indicates that the phosphorylated forms of PT3 and PT5 fit well into the PLP-binding site of plasmodial ornithine decarboxylase (PfODC), the key enzyme of polyamine synthesis, consistent with the ability to abolish ODC activity in vitro. Furthermore, the antiplasmodial effect of PT3 is directly linked to the capability of Plasmodium to trap this pyridoxyl analog, as shown by an increased sensitivity of parasites overexpressing PfPdxK in their cytosol, as visualized by GFP fluorescence.