Objective We assessed the 2-year course of Medically Unexplained Physical Symptoms (MUPS) in terms of symptom severity, physical functioning and mental functioning. Methods We used data of the PROSPECTS cohort study, consisting of 325 MUPS patients from several health care settings. Symptom severity (PHQ-15), physical functioning (RAND 36 PCS) and mental functioning (RAND 36 MCS) were assessed at baseline and 6, 12 and 24 months afterwards. We used Latent Class Growth Modeling (LCGM) to identify different course types, but as clinical usability of results was limited, we also analysed change scores and directions of change. Results LCGM identified three course trajectories for all outcomes: a “stable severe”, a “stable moderate” and an “improvement” trajectory. The greater majority of participants was assigned to the stable trajectories. However, we found high levels of within-trajectory heterogeneity. Based on total change scores (using only two measurement moments), physical symptoms of 27% of the participants deteriorated, while 63% improved. Based on Minimal Clinically Important Differences these proportions were 5% and 24%. Analyses of directions of change (using all measurement moments) showed that for all outcomes almost 80% reported a fluctuating course type, with clinically important fluctuations in 35-61% of participants. Conclusion Fluctuations in the course of MUPS were highly prevalent, but as fluctuations showed high interpersonal heterogeneity, these were not detected by LCGM. Improvement and deterioration rates based on change scores are in line with literature. However, based on the highly prevalent fluctuations, we conclude that temporal stability of these outcomes is limited.