This article is intended to establish the variables that explain the behavior of the CiteScore metrics from 2014 to 2016, for journals indexed in Scopus in 2017. With this purpose, journals with a CiteScore value greater than 11 were selected in any of the periods, that is to say, 133 journals. For the data analysis, a model of standard corrected errors for panel was used, from which a coefficient of determination of 77% was obtained. From the results, it was possible to state that journals of arts and humanities; business; administration and accounting; economics, econometrics, and finance; immunology and microbiology; medicine and social sciences, have the greatest impact.