This paper presents data on the measurement of aortic flow (Qa) using an electromagnetic flowmeter and the measurement of pulmonary blood flow (Qep) using a single-breath technique employing freon-22 as the soluble marker gas. The purpose was to assess the potential of a single-breath technique (using freon) as an effective way to estimate cardiac output (Q) non-invasively. The electromagnetic flowmeter was calibrated in vitro in an isolated heart-aorta preparation by collection of the ejected blood volume, thereby giving an absolute calibration of the Qa measurement. Using anaesthetized beagles Qa was measured immediately preceding (Qa,mean) and during (Qa,inst) measurement of Qep. These measurements were made before and after I.V. bolus doses of nifedipine used to increase cardiac output so as to measure a range of flows. The Qa was disturbed by the respiratory manoeuvre imposed as part of the single-breath technique, rising, falling and rising again as the lung volume was reduced, increasing with the test gas and reducing again, typically giving Qep = 1.39 +/- 0.13 l min-1, Qa,mean = 1.35 +/- 0.15 l min-1 and Qa,inst = 1.33 +/- 0.17 l min-1 (+/- S.D., n = 7 in each case). The standard deviations include interanimal variation as well as technique errors. Regression analysis of Qep on Qa,mean gave the highest correlation (r = 0.75) and the lowest coefficient of variation (11.7%), the values being 0.69 and 14.0% respectively for Qep on Qa,inst. These compare favourably with equivalent values for alternative methods reported in the literature, the strict calibration giving confidence that the single-breath technique provides data closely related to the aortic flow found prior to the respiratory manoeuvre, i.e. the undisturbed cardiac output.