Game is a human innate, pleasant and free activity, and is, as a multifunctional, non-specialized activity, associated with cognitive development. Therefore, it is one of the methods and strategies of the modern teaching process. The game makes teaching more interesting, motivates students, requires students to actively participate in teaching, which allows them better attention and concentration and easier and faster mastering of teaching content, which can sometimes be, especially for younger primary school students, difficult and uninteresting. Such are often the contents of grammar and spelling in the teaching of the Croatian language, which is the most comprehensive subject in primary education. As digital culture covers a large part of today's students' lives, it is good to implement computer games in the classroom, as an environment familiar to children in which they will enjoy, but at the same time learn. For the purpose of this paper, a research was set at one elementary school in Zagreb. The aim of the research was to examine students' attitudes about play in teaching and the use of computer games in teaching, whether the implementation of computer language game affects the acquisition of teaching content by students and whether gender of students affects the results of research. Students were divided into control and experimental groups. In the experimental group, motivation was conducted in the form of a computer language game for a class of practice and repetition of word breakdown at the end of a line, while in the control group it was not. After the (remaining) class, both groups solved a questionnaire created for the purpose of the research. The experimental group also solved a questionnaire related to computer language game. The data were analyzed in the SPSS program. The results showed that students have a positive attitude about the game in the classroom and about the computer game in the classroom, and that they would like to play more language games on the computer in the future. There was no statistically significant difference between the control and experimental groups in the implementation of the computer language game, nor between the sexes.