A best evidence topic in cardiothoracic surgery was written according to a structured protocol. The question addressed was whether videoassisted mediastinoscopy (VAM) is a more effective procedure than conventional mediastinoscopy (CM). A total of 108 papers were identified using the search as discussed below. Of which, eight papers presented the best evidence to answer the clinical question as they included a sufficient number of patients to reach conclusions regarding the issues of interest for this review. Complications, complication rates, number of lymph nodes biopsies, number of stations sampled and training opportunities were included in the assessment. The author, journal, date and country of publication, patient group studied, study type, relevant outcomes, results and study weaknesses of the papers are tabulated. Literature search revealed that CM is a safe procedure associated with low mortality (0-0.05%) and morbidity (0-5.3%). CM has high levels of accuracy (83.8-97.2%) and negative predictive value (81-95.7%). Training in CM can be difficult as the limited vision means that the trainer cannot monitor directly the dissection and the areas biopsied by the trainee as one operator and effectively see at any time. VAM is also a safe procedure with comparable results to that of CM in term of mortality (0%), morbidity (0.83-2.9%), accuracy (87.9-98.9%) and negative predictive values (83-98.6%). The main advantage is higher number of biospsies taken (VAM, 6-8.5; CM, 5-7.13) and number of mediastinal lymph node stations sampled (VAM, 1.9-3.6; CM, 2.6-2.98). VAM can be associated with more aggressive dissecting and that can lead to more complications. The use of VAM can provide a better and safer training opportunity since both trainer and trainee can share the magnified image on the monitor. All studies available are comparing heterogeneous groups of non-matched group of patients which can bias the outcomes reported. There is a lack of comprehensive randomized studies to compare both procedures and to support any preference towards VAM over CM. We conclude that there is actually very little objective evidence of VAM superiority over CM.