Poly(ADP-ribose) glycohydrolase (PARG) is being considered as a therapeutic target for the prevention of neurodegeneration. Here, we assessed the pharmacological tools available for target validation. The tannic acid derivative gallotannin inhibited PARG in a cell-free assay but had no detectable effect on PARG function in intact cells. Its cytoprotective actions were associated rather with the radical-scavenging potential of the compound. In astrocytes exposed to high concentrations of the nonoxidative DNA-damaging agent N-methyl-N'-nitro-N-nitrosoguanidine (MNNG), Poly(ADP-ribose) polymerase (PARP) inhibitors were fully protective, while gallotannin enhanced the damage. The compound N-bis-(3-phenyl-propyl)9-oxo-fluorene-2,7-diamide (GPI 16552), considered a potentially specific PARG inhibitor, had no effect in the different astrocyte death models compared with PARP inhibitors. In an in vitro PARG activity assay, the maximal inhibition that could be achieved with GPI 16552 was only 40% at a drug concentration of 80 microM. We conclude that neither GPI 16552 nor gallotannin are suitable for the evaluation of PARG in cellular death models, and that previous conclusions drawn from the use of these compounds should be interpreted with caution.