Problem solving is a complex cognitive activity that involves the construction of sequences of actions to reach a given goal. One powerful strategy is to identify analogies between the problem at hand and previously encountered ones. Relevant similarities between problems can be detected more easily if there is a high resemblance on the surface or with respect to structure. Earlier theoretical observations and performance data have pointed to two distinct kinds of analogical reasoning, direct solution transfer (transformational analogy) and the creation of a new solution based on adapted past reasoning processes (derivational analogy). In order to gain insights about the cognitive processes involved, we extend this work by an analysis of different kinds of verbal data. Planning protocols were collected prior to problem solving, and retrospective reports, evaluations, and instructions were elicited after the task was completed. Results show that the different kinds of analogical reasoning involved different degrees of analogy awareness, as reflected by the verbalizations. Derivational analogy involved problem solving on a more detailed and structured step-by-step basis than the more superficial transformational strategy, in which a simple matching procedure was employed.