As a response to troubling doubts about the success of Web-delivered courses that include problem solving and critical thinking skills, the authors gathered data from two online sections of the CIS Application Programming with Visual Basic course at Metropolitan State College of Denver over two semesters and compared it with the data from classroom sections. T-tests showed a significant difference in the means of student projects and tests that required application of theory learned. All sections of the course used the same syllabus and assignments and were taught by the same instructor. Internet students scored significantly lower in theory portions of exams and in projects. The authors conclude that more study is needed, but design and delivery of problem solving courses via the Web needs careful attention. INTRODUCTION As we move into the 21 century, educators know that Web courses at institutions everywhere are here to stay. The last decade of the 20 century saw a mad rush to put college courses, and entire college curricula online so that students could earn an entire degree without setting foot on a campus. Initial evaluations and statistical analyses consistently pointed to equality between classroom and the Web courses. Time and again, no significant differences were found in final grades between Web and classroom delivery of courses. The general consensus was that Web courses were working, and that colleges were justified in forging ahead with plans to put entire curricula online. The first courses to be put online in Information Systems departments were generally the lower level elective courses and courses which were survey or retention-oriented courses. That is, the material of the first computer literacy course, the first information systems survey course, introduction to the Internet, productivity software courses, and self-paced courses was simpler to port to the Web environment. An average knowledge of instructional design and a crash course in Web page editing with Front Page®, or even Web CT® or other course authoring software, was enough to move these “to-do list” type of courses to the Web. The actual content of the course was not within the Web pages of the course, but in textbooks and assigned readings listed on the course Web pages. When faced with putting courses that involve problem solving and higher analytical reasoning on the Web, there was concern that the initial success of the first Internet courses would not be repeated. Was there something missing? If so, it should be identified and action could be taken. It was time to re-examine the first successes of Web course deployment, and move to examining online courses in more depth. Are all Web courses working? And if not, why?