Introduction and objectives There are various recommendations for physical activity (PA). However agreement between all of these measures has not been established. Furthermore, given the challenges of measuring PA there is interest in evaluating whether a measure of exercise performance can be used as a surrogate measure to identify who is likely to achieve the recommendations. Methods A total of 184 people with COPD were recruited, 128 of which had complete data for these analyses. Participants wore the SenseWear Armband for 7 consecutive days and all performed an incremental shuttle walk test (ISWT). We extracted moderate to vigorous physical activity (MVPA) in bouts of ≥10 min using a 3 metabolic equivalent (MET) threshold and an individually prescribed MET threshold (based on performance on the ISWT). Average daily step count and the physical activity level were also calculated. Results There was poor agreement between the four PA recommendations, with agreement on all four achieved in only 30 participants. People were least likely to be active using MVPA in ≥10 min bouts using 3 MET threshold (21.1% active), and most likely to be active using MVPA in ≥10 min bouts using an individually prescribed threshold (64.9% active). It was not possible to identify a threshold on the ISWT that would reliably predict those that achieved any of the four recommendations. Conclusion Agreement between various physical activity recommendations is poor. This should be considered when measuring and describing physical activity adherence. The ISWT cannot be used to reliably predict adherence to physical activity guidelines.