There are a number of clinical methods to measure cardiac output, ranging from direct intracardiac catheterisation to non-invasive measurement of the arterial pulse. Each method has advantages and drawbacks. Relative comparison is limited by the absence of a widely accepted "gold standard" measurement. Cardiac output can also be affected significantly by the phase of respiration – intra-thoracic pressure changes influence diastolic filling and therefore cardiac output. This is especially important during mechanical ventilation, in which cardiac output can vary by up to 50% across a single respiratory cycle. Cardiac output should therefore be measured at evenly spaced points over a single cycle or averaged over several cycles.

Invasive methods are well accepted, but there is increasing evidence that these methods are neither accurate nor effective in guiding therapy. Consequently, the focus on development of non-invasive methods is growing.

Echocardiography is a non-invasive method of quantifying cardiac output using ultrasound. Two-dimensional (2D) ultrasound and Doppler measurements are used together to calculate cardiac output. 2D measurement of the diameter (d) of the aortic annulus allows calculation of the flow cross-sectional area (CSA), which is then multiplied by the VTI of the Doppler flow profile across the aortic valve to determine the flow volume per beat (stroke volume, SV). The result is then multiplied by the heart rate (HR) to obtain cardiac output. Although used in clinical medicine, it has a wide test-retest variability.It is said to require extensive training and skill, but the exact steps needed to achieve clinically adequate precision have never been disclosed. 2D measurement of the aortic valve diameter is one source of noise; others are beat-to-beat variation in stroke volume and subtle differences in probe position. An alternative that is not necessarily more reproducible is the measurement of the pulmonary valve to calculate right-sided CO. Although it is in wide general use, the technique is time consuming and is limited by the reproducibility of its component elements. In the manner used in clinical practice, precision of SV and CO is of the order of ±20%.[citation needed]