Parrots are popular companion animals but show prevalent and at times severe welfare issues. Nonetheless, there are no scientific tools available to assess parrot welfare. The aim of this systematic review was to identify valid and feasible outcome measures that could be used as welfare indicators for companion parrots. From 1848 peer-reviewed studies retrieved, 98 met our inclusion and exclusion criteria (e.g. experimental studies, captive parrots). For each outcome collected, validity was assessed based on the statistical significance reported by the authors, as other validity parameters were rarely available for evaluation. Feasibility was assigned by considering the need for specific instruments, veterinary-level expertise or handling the parrot. A total of 1512 outcomes were evaluated, of which 572 had a significant p-value and were considered feasible. These included changes in behaviour (e.g. activity level, social interactions, exploration), body measurements (e.g. body weight, plumage condition) and abnormal behaviours, amongst others. However, a high risk of bias undermined the internal validity of these outcomes. Moreover, a strong taxonomic bias, a predominance of studies on parrots in laboratories, and an underrepresentation of companion parrots jeopardized their external validity. These results provide a promising starting point for validating a set of welfare indicators in parrots.