AHP and Conjoint analysis methods are used to determine the priorities and preferences of groups or individuals in the decision-making process. These methods provide predictive results in many fields such as economics, politics, and environmental sciences. Multi-criteria decision-making methods have the potential to produce effective results in educational research where priorities and preferences, such as perceptions, attitudes, and beliefs, are in question. This study aims to determine the characteristics that teachers seek in a good mathematics problem and the features of problems they prefer in classroom practice, using AHP and Conjoint analysis. In this context, it is aimed to contribute to the literature in two areas. Firstly, to introduce AHP and Conjoint analysis, that are not yet widely used in educational research. Secondly, to examine the consistency between the characteristics sought in a good mathematics problem and the problems preferred in classroom practice. The study involved 35 mathematics teachers who were asked to provide data using pairwise comparison forms for AHP analysis and full profile cards for Conjoint analysis. The results indicate that there are differences between the priorities that teachers consider when defining a good problem and their problem preferences in classroom practice. According to AHP results, teachers determine the qualities of a good problem, functional features are ranked first, and physical features are ranked last. In contrast, when it comes to their problem preferences in classroom practice, solution-oriented features are ranked first, and physical features are ranked second through conjoint analysis results. This study demonstrates the applicability of AHP and conjoint analysis in educational research.