Background and objectives: To analyse the prevalence of positive prick-tests to all medicaments normally checked in allergy units when a patient is suspected of being allergic to anaesthetics. To establish the degree of agreement between the antecedents of a previous history of an allergic reaction to a medicament and the positive result, or not, to the specific prick-test for the said medicament. Methods: This was a prospective study, during 2003 and 2004, which analysed 473 patients referred by their doctors to allergy units to make retrospective diagnoses of an allergy to a drug. The prick-test was done using the undiluted drug. All patients were tested for 41 drugs. These include antibiotics, trimethoprim–sulphamethoxazole, non-steroidal anti-inflammatory drugs (NSAIDs) and perioperative drugs (PD): neuromuscular blocking drugs, latex, iodine, local anaesthetics, hypnotics, opioids and coadjuvants. Cohen's Kappa Index was used to determine the degree of agreement. Results: 71.5% of patients studied presented a positive prick-test. The largest number of positive cases was found in antibiotics (56.4%), followed by PD (15.6%), NSAIDs (14.4%) and trimethoprim–sulphamethoxazole (12.7%). Among PD, the highest prevalence of positive prick-tests was found for neuromuscular blocking drugs (5.3%). Agreement between the substance suspected of causing the allergic reaction and the positive prick-test was excellent for penicillin (Kappa = 0.74) and other antibiotics (Kappa = 0.721) and good for NSAIDs (Kappa = 0.47) and iodine (Kappa = 0.54). Conclusions: The prevalence of patients with positive prick-tests to PD occurred in 15.6% in this prospective cohort. Neuromuscular blocking drugs were found to have the highest prevalence of positive prick-tests. There is positive agreement when the substance responsible for the allergic reaction is suspected, otherwise agreement is low.