Polyelectrolyte (PE) wrapping of colloidal nanoparticles (NPs) is a standard method to control NP surface chemistry and charge. Because excess polyelectrolytes are usually employed in the surface modification process, it is critical to evaluate different purification strategies to obtain a clean final product and thus avoid ambiguities in the source of effects on biological systems. In this work, 4 nm diameter gold nanoparticles (AuNPs) were wrapped with 15 kDa poly(allylamine hydrochloride) (PAH), and three purification strategies were applied: (a) diafiltration or either (b) one round or (c) two rounds of centrifugation. The bacterial toxicity of each of these three PAH-AuNP samples was evaluated for the bacterium Shewanella oneidensis MR-1 and is quantitatively correlated with the amount of unbound PAH molecules in the AuNP suspensions, as judged by X-ray photoelectron spectroscopy, nuclear magnetic resonance experiments and quantification using fluorescent assay. Dialysis experiments show that, for a 15 kDa polyelectrolyte, a 50 kDa dialysis membrane is not sufficient to remove all PAH polymers. Together, these data showcase the importance of choosing a proper postsynthesis purification method for polyelectrolyte-wrapped NPs and reveal that apparent toxicity results may be due to unintended free wrapping agents such as polyelectrolytes.