The main goal of this paper was to identify malicious HTML code in web pages. For this purpose, the possibility of detecting inserted malicious HTML code using only HTML tags and their structure was investigated, and in general things about HTML documents and parsing HTML documents in general were explained. First of all, it is described how to extract HTML structure from HTML document using ready-made software for parsing HTML documents. After the HTML structure can be successfully extracted from the document, the algorithm for detecting newly inserted HTML structures of a document is explained. The latest version and the second latest version of the HTML document should be passed to the algorithm, and the algorithm will return all detected HTML structures as a result if it manages to detect any. Using this algorithm, an experiment was conducted on a small sample of documents. The method used in the experiment proved insufficient to achieve the goal of the experiment, and the possibility of detecting maliciously inserted HTML code using only HTML tags and their structure remains unexplored. For the method, improvements are proposed in this paper with which the method could be sufficient to investigate the problem. Another goal of this paper is to investigate the contribution of observing the HTML structure in solving the problem of recent duplicates in the web browser. There is already a solution to this problem, and this paper explains what an experiment that could determine whether observing the HTML structure is a potential alternative to an existing solution should look like.