BACKGROUND: Bibliometric and Altmetric analyses highlight key publications, which have been considered to be the most influential in their field. The hypothesis was that highly cited articles would correlate positively with levels of evidence and Altmetric scores (AS) and rank. METHODS: Surgery as a search term was entered into Thomson Reuter's Web of Science database to identify all English-language full articles. The 100 most cited articles were analysed by topic, journal, author, year, institution, and AS. RESULTS: By bibliometric criteria, eligible articles numbered 286,122 and the median (range) citation number was 574 (446-5746). The most cited article (Dindo et al.) classified surgical complications by severity score (5746 citations). Annals of Surgery published most articles and received most citations (26,457). The country and year with most publications were the USA (n = 50) and 1999 (n = 11). By Altmetric criteria, the article with the highest AS was by Bigelow et al. (AS = 53, hypothermia's role in cardiac surgery); Annals of Surgery published most articles, and the country and year with most publications were USA (n = 4) and 2007 (n = 3). Level-1-evidence articles numbered 13, but no correlation was found between evidence level and citation number (SCC 0.094, p = 0.352) or AS (SCC = 0.149, p = 0.244). Median AS was 0 (0-53), and in articles published after the year 2000, AS was associated with citation number (r = 0.461, p = 0.001) and citation rate index (r = 0.455, p = 0.002). AS was not associated with journal impact factor (r = 0.160, p = 0.118). CONCLUSION: Bibliometric and Altmetric analyses provide important but different perspectives regarding article impact, which are unrelated to evidence level.