Background: Existing research has indicated an association between funding source and publication of results that favour the sponsor’s product. These methodological reviews have focused mainly on the biased reporting of beneficial effects. However, potential bias with adverse effects remains uncertain as industry sponsors are required to present safety data that comply with regulatory requirements. Objectives: To provide an overview of methodological studies which have investigated the impact of source of funding in the publication and interpretation of adverse effects. Methods: To identify relevant studies, we searched ten bibliographic databases (including MEDLINE and EMBASE), handsearched key journals and conferences, contacted experts and undertook reference checking and citation searches. Two reviewers screened and selected studies which quantified the difference in reporting of adverse effects by source of funding. Data were extracted independently by two reviewers on the selection criteria used, the intervention, the adverse effects, the sources of funding of included studies, and the outcomes reported. Results: Six studies met the inclusion criteria. Five studies looked at only a single drug class, while one study evaluated a wide range of drugs. There were major differences between the studies in the methods used and outcomes assessed: (i) effect size: two studies reported a lower risk of harm favouring the sponsor’s product; (ii) reporting of statistically significant findings: two studies yielded conflicting results; one suggested industry sponsors were less likely to report significant harm for their product, while the other suggested more complete reporting by industry sponsors; (iii) subjective interpretation of data: two studies looked at the narrative reporting of adverse effects data and found that data may sometimes be interpreted in a way that favours the sponsor. Major methodological weaknesses in the available studies include the lack of baseline comparability between funded and unfunded studies, difficulty in determining funding source, and inconsistent outcome reporting. Conclusions: Some studies suggest that manufacturer-funded research tends to report the intervention as safer than non-manufacturer-funded research, but this was not always the case. The existing data is undermined by considerable methodological weaknesses, and higher quality research is needed.