1. The dermal wheal response to bradykinin is increased by drugs which inhibit angiotensin-converting enzyme, and thus provides a measure of angiotensin-converting enzyme activity at the tissue level. An insertion/deletion polymorphism of the angiotensin-converting enzyme gene predicts serum angiotensin-converting enzyme activity, but its relation to angiotensin-converting enzyme activity in tissue is unclear. 2. The relations between angiotensin-converting enzyme genotype and wheal responses to intradermal bradykinin were studied in 105 healthy subjects: 30 of genotype DD, 51 of genotype ID and 24 of genotype II. Dermal wheal area was measured by digitized planimetry and the angiotensin-converting enzyme genotype by polymerase chain reaction. 3. Bradykinin produced significant linear log dose-wheal area responses. The potency of bradykinin by parallel line bioassay did not differ significantly between the genotypes; the potency in the II subjects relative to DD subjects was 1.25 (95% confidence interval: 0.83-1.88). 4. Although the angiotensin-converting enzyme gene polymorphism is a consistent and powerful predictor of serum angiotensin-converting enzyme activity, it does not appear to predict tissue angiotensin-converting enzyme activity as measured by dermal responses to bradykinin.