The ability to solve problems using algorithms plays an increasingly important role in modern society, whereas programming, alongside with broadly understood digital skills, are considered to be one of the key competences of the future. The result of this trend is the modification in the Polish education system of the IT subjects’ core curriculum, under which teaching programming is planned at every stage of education. While teaching the skills of programming is important, it is not an easy task to achieve and hence it poses many methodological challenges. Researchers in this field are increasingly turning to new experimental methods, such as eye tracking techniques that allow to gain an insight into cognitive mechanisms and thus can provide objective information about the process of learning programming. The article discusses the results of authors’ own study, in which the state of declarative and procedural knowledge of students related to the forms of algorithm presentation was diagnosed. The questionnaires along with the tasks included in them, which the students solved in a traditional way and with the means of eye tracking techniques, were used in the study to track the process of solving comparable tasks presented on a computer monitor. The indicator of operational knowledge was the effectiveness of problem solving. The research was conducted on a group of 48 third-grade junior secondary school students. The obtained results (low level of correct answers) indicate that the situation in the area of learning the algorithmic skills of students requires improvement. The measurement data obtained using eye tracking allowed for an in-depth analysis and interpretation of visual activity of students. Therefore, it seems that eye tracking can be considered as a complementary research technique, enriching the state of knowledge on cognitive mechanisms that are triggered in the process of solving algorithmic problems.