Systematic reviews are secondary research projects that compile all the randomized trials addressing a particular question of interest. The processes and the procedures used to identify and select studies, extract and synthesize data must be standardized, transparent, and reproducible for a review to be classed as systematic. Well conducted, comprehensive systematic reviews of clinical trials provide the most valid evidence for directing the clinical practice in the light of evidence based medicine. However, systematic reviews rely on human endavour, and it is known that results of a systematic review of relatively smaller studies may be contradictory to the results of a later single large randomized controlled trial, or two systematic reviews addressing the same question may provide different answers. The key stages of systematic reviews and factors threatening the validity of results of systematic reviews have been discussed in this article.