Distribution of pulmonary blood flow has traditionally been regarded as determined by gravity. This view has been challenged recently by reports describing marked gravity-independent distribution of flow. These reports were based on experiments in which local blood flow was measured by methods that have not been thoroughly evaluated. In the present study, we showed that in the goat lung regional trapping of i.v. infused microspheres (O = 15 microns) correlated to endothelial uptake of a simultaneously i.v. infused diamine (r = 0.99, region size approximately 1.5 cm3, dry weight approximately 40 mg). This indicates that the deposition of microspheres reflects true regional pulmonary blood flow. Using the microsphere method, we found a marked gravity-independent heterogeneity in blood flow (coefficient of variation approximately 40%) in the awake goat. We could find no pattern related to anatomy that could account for this variability. We re-examined the influence of gravity by analysing the distribution of pulmonary blood flow in anaesthetized goats both in prone and supine positions. The dorsal to sternal distribution of flow appeared to be inverted when the animals were turned from prone to supine recumbency, indicating that gravity influenced the distribution of pulmonary blood flow along this axis. However, along the gravitational axis, distribution of blood flow varied considerably from lung to lung. It appears that in awake goats the distribution of pulmonary blood flow is the result of several different determinants.