Context: Previous research has shown a weak relationship between intended and actual training load in various sports. Due to variety in group and content, this relationship is expected to be even weaker during group rehabilitation. Objective: The goal of this study was to examine the relationship between intended and actual training load during sport-specific rehabilitation in a group setting. Design: Observational study. Setting: Three outdoor rehabilitation sessions. Participants: Nine amateur soccer players recovering from lower limb injury participated in the study (age 22 ± 3 y, height 179 ± 9 cm, body mass 75 ± 13 kg). Main Outcome Measures: We collected physiotherapists’ ratings of intended exertion (RIE) and players’ ratings of perceived exertion (RPE). Furthermore, Zephyr Bioharness 3 equipped with GPS-trackers provided heart rate and distance data. We computed heart rate–based training loads using Edwards’ method and a modified TRIMP. Results: Overall, we found weak correlations (N = 42) between RIE and RPE (r = 0.35), Edwards’ (r = 0.34), TRIMPMOD (r = 0.07), and distance (r = 0.26). Conclusions: In general, physiotherapists tended to underestimate training loads. To check whether intended training loads are met, it is thus recommended to monitor training loads during rehabilitation.