With the growing use of machine translation, more and more companies are also using post-editing services to make the machine-translated output correct, precise and fully understandable. Post-editing, which is distinct from translation and revision, is still a new activity for many translators. The lack of training, clear and consistent guidelines and international standards may cause difficulties in the transition from translation to post- editing. Aiming to gain a better understanding of these difficulties, this study investigates the impact of translation experience on post-editing performance, as well as differences and similarities in post-editing behaviours and trends between two languages of the same family (French and Brazilian Portuguese). The research data were gathered by means of individual sessions in which participants remotely connected to a computer and post-edited machine-translated segments from the IT domain, while all their edits and onscreen activities were recorded via screen-recording and keylogging programs. A mixed-methods approach was employed for the qualitative and quantitative analysis of the data. The findings suggest that there are no clear correlations between translation experience and post-editing performance, or post-editing experience and post-editing performance. However, other aspects such as the opinion regarding machine translation seem to be predictors of post-editing performance. Our analysis enabled us to combine multiple factors in order to identify the ‘best’ post-editors in our participant group. Finally, similar post-editing trends were observed for both target languages, suggesting that training, guidelines and automated aids could be targeted at language groups rather than at individual languages. The insight gathered will be useful for devising future post-editing guidelines and training programmes.