In comparing gas exchange responses of the methacholine- (MCh) challenged mongrel dog with leukotriene receptor blockers and placebo at different inspiratory O2 fractions (FIO2), we previously noted systematically different values of cardiac output as a function of drug administration and/or FIO2. This confounds identification of the effects of FIO2 and/or drugs on gas exchange, because shunt is well known to vary directly with cardiac output when other factors are equal. Accordingly, in six dogs we examined the dependence of combined shunt and low ventilation-perfusion (VA/Q) blood flow ("shunt") on cardiac output in the MCh-challenged mongrel dog. Two dogs breathed 100% O2, another two breathed room air, and the final pair breathed 12% O2 while cardiac output was altered several times by sequentially opening and closing arteriovenous fistulas every 10 min for approximately 90 min after a standard MCh challenge. On 100% O2, shunt increased by 11.0% of the cardiac output per 1-l/min increase in cardiac output. On room air, the value was 7.4%. With 12% O2 breathing shunt rose by only 2.2% per 1-l/min rise in blood flow. This FIO2 -dependent behavior of the shunt-cardiac output relationship was highly reproducible, both within and between animals. It suggests that the increase in shunt with cardiac output depends more on vascular tone of noninjured areas than on tone of the low VA/Q regions (which are hypoxic at all FIO2 values).