Background: Fatigue is a hugely common problem and symptom amongst patients in numerous chronic disease states. It negatively impacts patients’ quality of life and subsequently has a major social and economic burden. A crucial concept of better understanding fatigue is the perception of effort. Perception of effort can be measured using the Rating of Perceived Exertion (RPE) scale. Fatigue sufferers have a generally higher perception of effort. We can say that their world is a “higher effort world” whereby general tasks and activities require more effort. Reasons for this elevation of effort are not clear, however learning processes are suggested. If learning is part of the causes of elevated fatigue, then it is also likely that prediction of higher effort for actions extends to observation of actions i.e. ‘spill over’. Aims: Our first aim was to see if individuals can predict effort using RPE through observation alone. If this was possible, our second aim was to investigate how fatigue levels in patients influence this observational predictive capability and whether they have a prediction bias. Hypotheses: We propose that observer RPE (scores given by participants) and observed RPE (scores given by exercising subjects) will be strongly correlated in the patient and healthy participant groups. We also hypothesise that patients will have a significantly higher bias for observational RPE than healthy participants. Methods: Two online studies were designed whereby a patient group with fatigue symptoms (n=49) and a healthy group (n=74) were shown a series of videos of individuals with variable fitness levels exercising on a treadmill at different speeds (4km/h; 8km/h; 10km/h; 12km/h; 14km/h) and asked to rate the effort they felt the individuals were exerting using the Rating of Perceived Exertion (RPE) scale. The two groups were also shown an image of a flight of stairs and asked to rate the effort they feel they would need to exert using the RPE scale. Bias was calculated by subtracting observed RPE from the observer RPE. Results: Observer and observed RPE had a strong positive correlation amongst both healthy (rho(1848) = .74, p