Objective: Our goal was to analyze the effects of implementing problem-based learning (PBL) at the Azerbaijan University of Architecture and Construction. Method: We compared students that attended lectures and tutorials using PBL and students that attended lectures and tutorials using traditional teaching methods. In total, 737 students (intervention group) participated in PBL and 2201 in non-PBL (Control group A and Control group B) at the university. Descriptive statistics and a statistical Kruskal–Wallis test were used to analyze the data and determine whether groups could be compared. We used propensity score matching to make groups comparable and analyzed the differences between groups. Results: The results of the statistical analysis and descriptive statistical characteristics for the unmatched samples are presented in this paper. We also present the differences between the PBL group and two non-PBL groups based on the unmatched samples. All calculations were performed on one outcome measure: the final scores of students. The results show that even within a short period of time (one semester or academic year), applying PBL can have promising effects, although these effects are generally not statistically significant. However, after matching, the differences between the PBL and non-PBL groups were significant. Discussion and conclusions: Even a preliminary and limited introduction of PBL appears to have promising impacts on learning outcomes, as demonstrated by students’ final scores on subjects that were taught using the PBL approach. The results suggest that post-Soviet countries with education traditions that differ from Western ones can benefit from PBL. It is advised to prolong the implementation of PBL and advocate for wider dissemination of this method in Azerbaijan and elsewhere in post-Soviet countries.