Although adequate assessment of pain and anxiety during burn wound care serves important clinical and scientific goals (e.g., determination of medication dosage and evaluation of treatment effects), few data are actually available. Studies which compare self-reported pain with observational ratings frequently suffer from small sample sizes or questionable data analysis techniques. This paper presents a study in which 126 burn wound dressing changes were independently rated by patient and nurse(s). Variables investigated were overall pain, worst pain, overall tension and tension just before the procedure. Also studied were the effects of age, extent of burn damage, medication use and nurses' experience with burn patients. It is concluded that although there are significant relationships between patients' and nurses' ratings, these relationships are not very strong. In contrast to self-reports, observational ratings seem to be affected by patient's age. Patients' ratings of tension correlate with the amount of analgesic medication used. It is argued that it is not useful to discuss the present and earlier studies only in terms of correctness or incorrectness of observational ratings. Recommendations for future studies include the study of pain-related behaviors, coping mechanisms and effects of treatments. Considering the vast differences in prescription regimes among centers, a multicenter trial would be particularly interesting.