The cause of category III A prostatitis, chronic prostatitis/chronic male pelvic pain syndrome category A (CP/CPPS A), is uncertain. Treatments for it are based on consensus opinion rather than on scientific data. Our aim was to examine the effect of zafirlukast, a leucotriene antagonist, on the symptoms of CP/CPPS A in our genitourinary (GU) medicine unit. CP/CPPS A was diagnosed by comparative white cell counts of split urine (Stamey) analysis or by finding an excess of polymorphs in expressed prostatic fluid. Symptom change was assessed by the National Institutes of Health Chronic Prostatitis Symptom Index (CPSI). Patients were given zafirlukast or placebo for four weeks in a random double-blind fashion. All patients also received doxycycline. In all, 31 patients were asked to participate and 17 entered the study. No difference in outcome could be shown between the active (10) and placebo (seven) patients. Zafirlukast cannot be demonstrated to be useful in the symptomatic treatment of CP/CPPS A. The problems of recruitment into this study (in spite of a large number of patients with prostatic type pain being seen in our unit) suggest that multicentre treatment trials using non-invasive diagnostic techniques such as the CPSI (rather than single GU medicine units diagnosing CP/CPPS A by uncomfortable direct prostatic testing) are likely to be the most effective and objective methods of undertaking treatment trials in the CP/ CPPS A field in the future.