In the last 10 years, health sciences literature has seen a significant increase in papers that report findings of meta-analyses. Meta-analysis overcomes the weakness of conventional narrative literature review. Moreover, it has established itself as a significant tool in clinical research and evidence-based medicine. But despite its station in the field, it is not without controversy. The aim of this paper is to present the main controversies that surround meta-analyses and to discuss the most important issues of conduct and reporting of results derived from these studies. Controversial issues include the ability of the investigators to combine studies that differ in important aspects; such as study populations, experimental designs and quality controls, and the like. It has been argued that selection bias could exist. Thus, it is common for the results of metaanalyses to be contradictory. As any other statistical procedure or analytical approach, meta-analyses too can be misused or abused. Misleading results could be avoided if certain basic principles are followed in the conduct of a meta-analysis. In particular, attention should be given to: (a) the methods of literature search for the studies to be included in the analysis, (b) the measures taken to reduce or eliminate bias, (c) the inclusion and exclusion criteria of the individual studies from the analysis, (d) how the data will be extracted from the literature, and (e) how the data will be analyzed statistically. Reports of meta-analytical studies carried out with the above parameters in mind yield a more objective appraisal of the evidence, and a more precise estimate of the treatment effect.