This study aimed to examine whether a graphical abstract (GA) on a publisher's official website affected an article's usage and citations. Articles published in Molecules during 2016 (n = 1389) and 2017 (n = 1804) were selected as the data sets. Propensity score matching analysis was conducted to examine the data sets. The results showed that articles with GAs had significantly greater text abstract usage than those without GAs. There were no significant differences between the two groups (articles with/without GAs) on the usage of the full text as well as citations, both in the data set of 2016 and the data set of 2017. Our study concluded that a GA played a role in attracting attention with ‘the first impressions’ to gain more clicks of text abstract for an article, however, it took no advantage in receiving more usage of full text and citations of the article.