This paper examines the definition of a problem and at the process of problem solving. An analysis of a number of first and third year chemistry examination papers from English universities revealed that over ninety per cent of the "problems" fell into the "algorithm" category. Using Bloom's taxonomy and the same examination papers, we found that the categories of knowledge and comprehension were predominating. At the Open University we have tried to find out how students go about the solving of more unfamiliar style problems. There was a greater tendency for males to get into the problem more quickly, but there seemed to be no gender difference in overall performance, suggesting a male tendency to start without allowing time for due consideration of the task. In contrast, a larger proportion of the female population spent time thinking before embarking on a particular route. Finally, the type of work carried out in the chemistry laboratory is considered, where more formal activities of the verification of known laws and effects etc. dominate. It is good that new findings from research are introduced into the undergraduate curriculum, but there is also a need for removing redundant material. (Contains 4 tables, 6 figures, and 1 note.)