We tested the hypotheses that angiotensin-converting enzyme insertion/deletion (I/D) and angiotensinogen 235 methionine/threonine (M/T) substitution gene polymorphisms influence angiotensin-converting enzyme and angiotensinogen serum concentrations and cardiac dimensions in 91 monozygotic and 41 dizygotic twin pairs. Cardiac dimensions were determined echocardiographically. Angiotensin-converting enzyme levels were 24 +/- 11, 43 +/- 18, and 58 +/- 24 U/L for the II, ID, and DD genotypes, respectively (P < .01). Posterior wall thickness was 8.1 +/- 1.3, 8.6 +/- 1.7, and 8.9 +/- 1.9 mm for these genotypes (P < .05). Angiotensin-converting enzyme levels were correlated with posterior wall thickness (r = .15, P < .05). The intrapair differences in angiotensin converting enzyme levels for monozygotic, concordant dizygotic, and discordant dizygotic twins were 1.36 +/- 1.6, 1.86 +/- 1.6, and 17.25 +/- 4.3 U/L, respectively. The angiotensinogen M/T genotypes exerted no influence on cardiac dimensions or on angiotensinogen concentrations. The additive genetic effect on angiotensin-converting enzyme levels (0.49), on posterior wall thickness (0.26), and on septum thickness (0.37) was significant (P < .01), although shared and nonshared environmental effects were also identified. Our data confirm the impressive effect that the angiotensin-converting enzyme D allele exerts on angiotensin-converting enzyme plasma levels. Furthermore, our data also suggest that the angiotensin-converting enzyme gene locus is primarily responsible for angiotensin-converting enzyme plasma levels. Our twin study also indicates that the angiotensin-converting enzyme gene locus is genetically linked to posterior wall thickness. The correlation between angiotensin-converting enzyme levels and posterior wall thickness suggests that this effect is exerted by angiotensin-converting enzyme. We were unable to demonstrate genetic linkage between the angiotensinogen gene locus and cardiac dimensions in this study. (Hypertension. 1997;29[part 2]:165-170.)