This contribution aimed at developing an understanding of student teachers' conceptions of guided discovery teaching approaches. A cross-sectional survey design involving eleven secondary mathematics teachers who had enrolled for an in-service mathematics education degree was used to address the research question: What are undergraduate student teachers' conceptions of deductive and inductive teaching approaches? Task-based interviews were used in conjunction with oral interviews as settings for unravelling students' conceptions of the two teaching approaches. Drawing in part from Ausbel's learning theory and Tall's notion of a met-before, the study also aimed at assessing the students' level of grasp of fundamental limitation of empirical explorations despite many benefits associated with them such as helping in identifying patterns, use in formulation and communicating of conjecture, and providing insights on what needs to be solved. Verbatim transcriptions from follow up interviews and textual data from task based interviews were subjected to directed content analysis to infer meaning about students' conceptions of guided teaching approaches. Qualitative data analysis using in part Robert Moore's notion of concept usage uncovered conceptual limitations that include inconsistencies in student teachers' definitions of the teaching approaches, use of specific examples instead of arbitrary mathematical objects in illustrating analytic teaching. Limitations identified should be given attention by mathematics educators in order to increase understanding of the approaches among teachers. Research studies into factors contributing to limitations identified can go a long way in improving the teaching and learning of school mathematics.