This paper, based on the inquiries of the General Mathematics Project, makes a case for naturalistic research in mathematics classrooms. In explicating the case, consideration is given to the need for information on classroom practices and the consequence of those practices on teaching and learning. Also considered is the need for research that addresses the problems of practice from the practitioner's perspective. Included is a discussion of the General Mathematics Project's evolution and tentative results. Some of those results are that teachers do teach general-math students differently from algebra students and that most teachers have unusual difficulty teaching general math and often feel only marginally or not at all successful. (Author/MP)