There are now two experimental approaches to capillary permeability: whole-organ studies and single-capillary experiments. It is not possible to obtain permeability coefficients in the true sense of the word from whole-organ experiments since the capillary surface area is unknown. Other limitations of whole-organ approaches are discussed. Results from 'osmotic transient' experiments are compared with those from 'indicator diffusion' experiments. 'Osmotic transient' experiments give higher values for capillary diffusion capacity than 'indicator diffusion' experiments. Reasonable agreement is obtained when the reflection coefficient is taken into account. Thus far, single-capillary experiments have only provided results from capillaries in the frog mesentery, where the potassium permeability was about 70 X 10(-5) cm . sec-1, or 10 times the value from whole-organ experiments on muscle. The discrepancy may reflect true differences between capillaries in the two tissues or may be caused by methodological limitations. There is a need for comparative studies of permeability with the whole-organ and the single-capillary technique applied to the same tissue.