Claustrophobia is associated with cardiovascular magnetic resonance (CMR) due to the nature of the technique and lengthy examination times. We report the incidence and treatment of claustrophobia in a prospective study on the use of intravenous (IV) and oral diazepam for CMR from August 1997 to June 1999 in 1754 adult patients referred for clinical (1226) and research (528) purposes. Patients under 16 yr of age and adults with learning difficulties were excluded. The initial refusal rate of clinical and research patients was 4.2% (54 clinical and 19 research). Because ethical approval to administer diazepam to research patients had not been granted at the beginning of this study, further analysis of the results excludes the 19 research patients. Of the 54 clinical patients, 31 were given IV diazepam (mean dose, 7.5 mg; range, 2.5-20 mg) with successful scanning in 30 (97%). Eight patients refused sedation, and scanning was not possible. The examination was attempted without sedation by five patients in the claustrophobic group and was terminated early; however, sufficient diagnostic information had been acquired in all cases. A further four patients took oral diazepam up to 1 hr before their appointment, with a 100% success rate. Three patients refused to attend the department due to known severe claustrophobia (2) and concerns over gradient noise (1), and three had medical contraindications to diazepam. Therefore, after the administration of diazepam to alleviate claustrophobia, the failure rate decreased to from 54 (4.4% of clinical patients) to 20 (1.6%) patients, a reduction of 63%. When it is possible to give diazepam to patients with claustrophobia at the time of CMR, it is a safe, predictable, and highly effective method of obtaining a successful result. A protocolfor the use of diazepam has been developed and is now also approved by our institutional ethics committee for use in research patients.