Obesity is regarded as a pro‐inflammatory state. It is associated with low circulating levels of the adipokine, adiponectin, which is considered to be an anti‐inflammatory. However, adiponectin knockout mice do not consistently demonstrate pro‐inflammatory phenotypes, suggesting more complexity in the in vivo immunomodulatory effects of adiponectin than originally anticipated. Moreover, adiponectin exerts pro‐inflammatory effects in some experimental systems. This contradiction has been resolved by hypothesizing that adiponectin induces tolerance to inflammatory stimuli, notably Toll‐like receptor (TLR) ligands. We noticed that this effect resembled lipopolysaccharide (LPS) tolerance and therefore tested adiponectin from a variety of sources for LPS contamination. All adiponectin tested carried low levels of LPS in the range of 1–30 pg/μg of adiponectin, sufficient to produce final LPS concentrations in the pg/ml range under experimental conditions. We found that induction of tolerance to TLR ligands by adiponectin in human monocyte‐derived macrophages could be reproduced by such LPS concentrations. Moreover, the LPS antagonist, polymixin B, substantially inhibited induction of tolerance by adiponectin. Furthermore, polymixin B and a naturally occurring antagonist LPS were able to partially attenuate induction of tumour necrosis factor‐α and interleukin‐6 in human monocyte‐derived macrophages by adiponectin. Polymixin B also inhibited nuclear factor‐κB and mitogen‐activated protein kinase signalling elicited by adiponectin. We therefore propose that some of adiponectin’s immunomodulatory effects, in particular, its TLR‐tolerising actions in human monocyte‐derived macrophages, may be confounded by induction of tolerance by contaminating LPS.