Both students and teachers need different competences for scientific reasoning. Apart from the required content knowledge and the ability of using it adequately, both groups need elaborated knowledge of strategies for decision-making and argumentation. These competences concerning decision-making are highly dependent on how frequently students are given the chance to argue in science classes. This article pools the results of three different research projects in chemistry education which focus on these three aspects: Students’ competences, the classroom situation with regard to communication and reasoning and teachers’ competences. Students’ skills needed for discussing chemistry content and for decision-making in scientific contexts are analyzed first. Following this, the opportunities for improving these skills during science lessons are researched and related to results from a study of teachers’ pedagogical content knowledge (PCK) in this domain. The analysis of students’ and teachers’ communication skills is conducted in two different ways, paper-pencil tests and video analysis. Paper-pencil tests are used to assess students’ performance in reasoning. The test items deal with chemistry-specific situations, including opportunities for decision-making, which are typical of socio-scientific issues (study 1). The study shows that students’ decision-making skills are poor when the topics deal with scientific contexts, but that students perform be tter when dealing with everyday-life contexts. One reason might be the lack of reasoning in chemistry lessons, as shown by a video study conducted in chemistry classes. Students’ and teachers’ in-class behavior and communication patterns are analyzed with regard to students’ and teachers’ contributions (study 2). The amount and the quality of students’ and teachers’ statements as well as the interactions, especially teachers’ reactions to students’ statements are investigated. Teachers’ way of negotiating and guiding classroom discussions should depend on their PCK in this specific field. Due to a lack of studies on correlations between teachers’ PCK and classroom activities, the third study focuses on the development of adequate tests on PCK. Tests on content knowledge (CK) are developed in addition to prove the expected correlations between PCK and CK (Baumert & Kunter, 2006).