Formative Assessment (FA) is a promising teaching practice for supporting students' learning at school. However, implementing FA into teaching (mathematics) is challenging, effects of implementation of FA vary between empirical studies. Therefore, recent studies additionally consider instructional quality of teaching when analyzing effects of FA on students' learning. The current exploratory study builds on this idea and highlights the added value of considering both generic and subject-specific instructional quality (GIQ and SSIQ) when analyzing effects of FA: Based on a re-analyzation of data from the research project Co[superscript 2]CA, 856 students from 39 mathematics classes of German middle schools took part in an intervention control-study aiming at implementing FA into mathematics education at school. On the student level, students' mathematics achievement is assessed by standardized tests before and after the intervention. During the intervention, on the class level GIQ and SSIQ are assessed by low- and high-inference ratings of 72 video-taped lessons. GIQ is based on the model of Three Basic Dimensions, SSIQ is based on the normative idea that mathematics teaching should offer opportunities for a deep understanding of subject matter and for building up competencies. Multi-level regression analyses highlight different (interaction) effects of GIQ and SIQ on students' performances. It becomes obvious: Considering GIQ and SIQ can have an added value for the better understanding of implementing FA into teaching (in mathematics education).