After the start of heroin (diacetylmorphine)-assisted treatment to a selected group of chronic treatment-resistant heroin-dependent patients in the Netherlands,we reported about work-related eczema and positive patch tests to heroin in some nurses and nasal and respiratory complaints. To investigate the prevalence of heroin contact allergy,we started a questionnaire-based study with follow-up by allergological examinations. Of 120 questionnaires sent, 101 (84% ) was returned: 67 from nurses and 34 from other employees. Of 101 workers, 38 (38% ) had reported work-related complaints: 33 of 67 (49% ) nurses and 5 of 34 (15% ) other employees. Patch tests to heroin were performed in 24 nurses and were positive in 8 (33% ). All the 8 had eyelid or facial eczema and, in 6, accompanied by mucosal or respiratory complaints. The prevalence of heroin contact allergy in this study was 8% (8/101) among all employees and 12% (8/67) among nurses. Respiratory and mucosal complaints could not be ascribed to a contact allergy, and in these cases, serum was analysed for specific immunoglobulin E to heroin. A type 1 allergy to heroin could not be shown. These complaints are possibly due to the histamine-liberating effect of heroin, to atopic constitution, to a combination of these factors or less likely to other non-allergic factors. After the start of heroin (diacetylmorphine) -assisted treatment to a selected group of chronic treatment-resistant heroin-dependent patients in the Netherlands, we reported about work-related eczema and positive patch tests to heroin in some nurses and nasal and respiratory complaints. To investigate the prevalence of heroin contact allergy, we started a questionnaire-based study with follow-up by allergological examinations. Of 120 questionnaires sent, 101 (84%) was returned: 67 from nurses and 34 from other employees. Of 101 workers, Patch tests to heroin were performed in 24 nurses and were positive in 8 (33%). 38 (38%) had reported work-related complaints: 33 of 67 (49%) nurses and 5 of 34 (15%) other employees. All the 8 had eyelid or facial eczema and, in 6, accompanied by mucosal or respiratory complaints. The prevalence of heroin contact allergy in this study was 8% (8/101) among all employees and 12% (8/67) among nurses Respiratory and mucosal complaints could not be ascri bed to a contact allergy, and in these cases, serum was analyzed for specific immunoglobulin E to heroin. A type 1 allergy to heroin could not be shown. to a combination of these factors or less likely to other non-allergic factors.