This project investigates two young professional translators’ problem-solving and decision-making behaviour during revision processes. It sets out to qualitatively describe the complexity of interplay involved in problem-solving and decision-making in translation revision, using think-aloud protocols as a research method. The data I elicited suggest that, for a revision point to occur, the translator first has to find a translation problem. However, the translation problem itself can evolve over time in the revision process in either a divergent or convergent manner. In other words, a single translation problem can be subdivided into several smaller problems and be tackled individually. Meanwhile, the translator may choose to merge several problems into a single problem that requires a holistic problem-solving approach. In terms of decision-making, the translator does not generally verbalise his/her reasons for choosing a translation solution. Nevertheless, s/he has an appropriateness threshold in mind, so that s/he can judge and compare the appropriateness of translation choices and make a decision accordingly. A tentative model of end-revision problem-solving and decision-making has been produced to summarise the findings of this project.