Over the past several decades, empirical support for inquiry-based forms of science teaching has accumulated. Critical voices, however, also exist, claiming that minimally guided teaching approaches might be less effective than guided forms; they might even be harmful, particularly for novice students, due to the production of cognitive overload. In this study, a more guided type of inquiry-based forms of instruction was therefore selected, which can be traced back to Alfred N. Whitehead: the so-called learning cycle approach. This approach was applied to science classes with 280 socioeconomically disadvantaged students in lower secondary education at low-track schools in Austria. Using a pre-test post-test experimental-control group design, this study investigated the effects of this instructional approach not only concerning students' cognitive development but also concerning their state and trait emotions, their behavioral engagement in class, and their academic self-concept. The results indicated that the application of the learning cycle approach had a positive impact on most of the measured students' trait emotions and state emotions during the romance and generalization phases of a full learning cycle. It further improved the students' behavioral engagement during lessons and strengthened their academic self-concept. Finally, it was found that the learning cycle approach positively affected the students' cognitive development after the 2-year treatment. Suggestions for improving the implementation of this teaching approach in science classes are given; some proposals for optimizing future research on learning cycle teaching are also made.