Poor compliance may be responsible for symptomatic decompensation or neurohormonal "escape" in patients with heart failure treated over the long term with angiotensin-converting enzyme-1 (ACEI) drugs. Serum ACE activity is a poor index of neurohormonal suppression or haemodynamic effect after ACE-inhibitor treatment. Serum ACE activity may, however, be a useful index of compliance with treatment, as serum ACE is sensitive to the presence of an ACE inhibitor in the blood. Sixteen normotensive male volunteers of known ACE genotype received 7 days of randomised, double-blind therapy on four occasions 2 weeks apart with lisinopril 20 mg (L) or matched placebo (P) to simulate (A) noncompliance (all P), (B) full compliance (all L), (C) partial compliance (L days, 1, 3, 6; P days, 2, 4, 5, 7), or (D) single dose (L day 7; P, 1-6). Supine (30 min) blood pressure (BP)/heart rate (HR), ACE, and angiotensins were measured on d7 before dose and 4-6 h after dose. Results are mean +/- 1 SD. BP showed the expected small decrease with active treatment on d7 (B or D) but not with placebo (A) or partial compliance (C). Prestudy serum ACE, despite a wide range (16-124 U/L), was reproducible within subjects [coefficient of variation (CV), 1.7%]. Serum ACE activity, before (41.9 +/- 30) and after (41 +/- 30) angiotensin (A) I or II, were unaffected by treatment (placebo A). Active treatment (B) resulted in very low serum ACE activity and d7 and a small further suppression after dosing (before, 3.9 +/- 4; after, 1.8 +/- 4). AI was elevated in this group with further elevation after dosing (before, 234 +/- 116; after, 551 +/- 250). AII was only modestly reduced from baseline and showed little further suppression after dosing (before, 7.8 +/- 4; after, 6.3 +/- 5). Partial compliance (C) showed low ACE but no reduction after treatment (before, 7 +/- 3; after, 7 +/- 4), an elevated AI but no dosing effect (before, 187 +/- 198; after, 200 +/- 151) and reduced AII but with no further dose suppression (before, 6.4 +/- 3.4; after, 7 +/- 4) induced increase in peptide (compared with B). Single-dose treatment (D) showed ACE inhibition as expected (before, 47 +/- 30; after, 2.2 +/- 3). There was a dosing-related increase of AI but to a lesser extent than seen with chronic active dosing (B) (before, 39 +/- 10; after, 240 +/- 200). In contrast to long-term dosing, there was marked ANG II suppression (before, 8.8 +/- 4; after, 2.9 +/- 3). With this long-acting ACEI in a dose relevant to congestive heart failure management, we suggest that 4-6 h after-dosing serum ACE (< 5 EU/L) and elevated ANG I (> 300 pg/ml) can be used to confirm compliance with treatment. These absolute values may be altered in patients treated concomitant with loop diuretics. In principle, however, this may be a useful tool in clinical trials or in clinical practice after further work has been done to assess the limits in patients across the doses and across the range of available drugs used.