The existence of a dose-response relation between nitrous oxide concentration and regional dysfunction in compromised myocardium, and whether or not halothane-induced myocardial depression alleviated this regional dysfunction was examined. Nitrous oxide was administered to eight dogs with experimentally induced left anterior descending coronary artery (LAD) critical stenosis during fentanyl (100 micrograms/kg bolus plus 1.5 micrograms.kg-1.min-1) anesthesia. Two modes of nitrous oxide administration were employed: gradual (in steps of 20% inspired, i.e., 0%, 20%, 40%, and 60% inspired) and abrupt (0-60% inspired). Regional myocardial function was assessed by sonomicrometry. Regional dysfunction in the compromised myocardium, in the form of postsystolic shortening (PSS), increased above baseline levels during 40% (4.2 +/- 2.3% to 12.1 +/- 3.9%, P less than 0.05) and 60% (4.2 +/- 2.3% to 12.5 +/- 3.6%, P less than 0.05) inspired nitrous oxide (gradual administration) and also during abrupt 60% nitrous oxide administration (6.4 +/- 2.6% to 9.9 +/- 3.2%, P less than 0.05). After abrupt 60% inspired nitrous oxide administration, halothane (0.7% inspired) was introduced and caused decreases in mean arterial pressure (106.1 +/- 4.5 mm Hg to 76.2 +/- 5.5 mm Hg, P less than 0.05) and peak LV dP/dt (1700 +/- 150 mm Hg/sec to 1100 +/- 100 mm Hg/sec, P less than 0.05). Halothane caused a marked increase in PSS (9.9 +/- 3.2% to 30.8 +/- 12.6%, P less than 0.05). Thus nitrous oxide administration caused regional dysfunction in myocardium supplied by a critically narrowed LAD whether administered gradually or abruptly and at concentrations as low as 40% inspired.(ABSTRACT TRUNCATED AT 250 WORDS)