Background: Objective recognition of pain in horses has been studied extensively, however studies on objective pain assessment in donkeys are limited, and the available scales are not validated. Objectives: This study describes scale construction and clinical applicability of a Composite Pain Scale (do-CPS) and a Facial Assessment of Pain scale (do-FAP) for acute pain in donkeys. Study design: observational. Methods: The study included 159 adult donkeys (n = 44 patients, n = 115 control donkeys) which were directly observed at The Donkey Sanctuary. Patients were presented with lameness (24), colic (7), head related pain (7) or post-operative pain (6). Based on equine scales specific potential elements and scores for donkeys were developed in a pilot study. The observers were not involved in donkeys’ clinical management. For each animal, the score of each element in both scales was assessed by two groups of independent observers. When applicable the patients were followed over time, once or twice daily. Patients and control groups were compared by Mann Whitney-U-tests. Results: The inter-observer reliability was strong for do-CPS (R2 = 0.95, p<0.001) and good for do-FAP (R2 = 0.77, p<0.001). Patients had significantly higher pain scores, compared to control donkeys (p<0.001 for both do-FAP and do-CPS). Sensitivity overall for the do-CPS (73%), do-FAP (68%), and specificity do-CPS (99%), do-FAP (75%) were good. Sensitivity and specificity for “lameness” were strong in do-CPS (92% and 100%, respectively). Sensitivity and specificity for “colic” were strong for both do-CPS (71% and 100%, respectively), do-FAP (95% and 79%, respectively). Main limitations: Observers could not be masked to the patients’ condition. More patients are needed with painful conditions other than lameness. These scales will be validated in a planned follow-up research. Conclusions: Objective pain assessment in donkeys is possible and may support objective evaluation of treatment of donkeys with acute pain.