Objective: In adults hospitalized with an acute or chronic respiratory condition, to determine what has been reported regarding exercise programmes in terms of content, tolerability, evaluation and adverse events. Data sources: A systematic search was conducted of electronic databases (PubMed, EMBASE, CINAHL, PEDro, The Cochrane Library), trial registries and conference abstracts (Thoracic Society of Australia and New Zealand Annual Scientific Meeting, the European Respiratory Society Congress, the American Thoracic Society International Conference). Review methods: Studies were included if they (1) recruited adults hospitalized with an acute or chronic respiratory condition, (2) described an exercise programme that targeted peripheral muscles and (3) reported that ⩾80% of the sample had initiated training within 72 hours of hospitalization. Results: The last search was conducted on 2 June 2019. Of the 6282 records identified, 20 met the study criteria. These described 18 separate studies (2018 participants). Studies were conducted in adults hospitalized with an exacerbation of chronic obstructive pulmonary disease or with community-acquired pneumonia. The content of exercise programmes included aerobic and/or resistance training, neuromuscular electrical stimulation, whole-body vibration or movement out of bed. In eight studies (44%), the initial session was prescribed using objective measures of exercise capacity, peripheral muscle force and the ability to undertake activities of daily living. Across 7420 training sessions, seven adverse events were reported. Conclusion: Methods used to prescribe and titrate exercise programmes in adults hospitalized with an acute or an exacerbation of a chronic respiratory condition were disparate. When reported, programmes were well tolerated and adverse events were infrequent.