Background: Diagnosis of complex regional pain syndrome type I (CRPS I) is based on clinical observation of symptoms. As little information is available on the reliability of CRPS I diagnosis, we evaluated the agreement between therapists with regard to the presence and severity of CRPS I and its symptoms. Methods: The interrater reliability was evaluated in 37 presumed CRPS I patients by three observers; one consultant anesthesiologist and two resident anesthesiologists. Patients were assessed on the basis of Veldman's CRPS criteria. Results: The interrater reliability for diagnosing CRPS I was good for the majority of observer combinations. The percentage of agreement for the absence or presence of CRPS I was good (88%-100%). Cohen's Kappa's ranged from 0.60 to 0.86. The agreement for the mean symptom score ranged from 70.2% to 88.6%; Kappa's were lower and showed more variation. Interrater reliability for assessment of the severity of CRPS I and its symptoms was poor. Factors influencing the interrater reliability were symptom type, individual observers and sample population. Conclusion: Diagnosing CRPS I can be performed on the basis of clinical observation. Further assessment of severity of CRPS I and its symptoms should be performed with reliable and valid measurement instruments.