This paper presents and analyzes the results of the application of computer-adaptive testing in knowledge evaluation of the programming language C++. The participants of the research were college students majoring in computer engineering from a polytechnic college (Republic of Serbia) with an average age of 20. The research was conducted over a two-year period. Overall 199 students took part in this research. Two groups of students, the experimental, and the control one were observed. The students in the experimental group took the computer adaptive test and the students from the control group did the paper-and-pencil test. For the purposes of determining the effects of applying the computer adaptive test for knowledge evaluation, the adaptive test was realized in the MATLAB software package. The test consists of multiple choice questions with five possible answers. The questions for the test are divided into three clusters according to the level of difficulty (easy, medium and hard). The score is calculated from the level of difficulty. The difficult questions earn more points than the easy ones. The examinees may answer the same percentage of questions correctly, but an examinee will gain a better score assuming they correctly answered more difficult items. The results of the research show that there is a statistically significant difference between the results of the control and experimental group. The students who worked on a computer-adaptive test achieved a higher average score than the students who did the traditional test.