The Task-Based Approach (TBA) has gained popularity in the field of language teaching since the last decade of the 20th Century and significant scholars have joined the discussion and increased the amount of analytical studies on the issue. Nevertheless experimental research is poor, and the tendency of some of the scholars is nowadays shifting towards a more tempered and moderate stand on their claims. Reasons for that are various: the difficulty in the implementation of the method in the classroom, the difficulty in elaborating materials following the TBA and the scarcity of task-based manuals count as important and perhaps decisive arguments. But there are also theoretical implications in the TBA which do not seem to be fully convincing or may lack sound foundations. In this paper I will attempt to describe the TBA critically, pointing out what I consider positive in this approach, and underlining the inadequacy of some assumptions and conclusions. The design of a new TBA model is not the goal of this study. But the conclusions suggest that tasks may contribute to the production of a more refined and complete foreign language syllabus, helping to motivate the students and focus the attention of teachers and learners on meaning and communicative language use.