Catecholamines in plasma may be measured to assess sympathoadrenal activity. Numerous assay methodologies have been published, illustrating the fact that there are many analytical problems. Different methodologies are discussed briefly. A plea for better validation, especially with regard to specificity (which should not be confused with sensitivity or reproducibility), is made. Plasma NA is a frequently used marker for sympathetic nerve activity in humans, but the data obtained are often misinterpreted due to lack of appreciation of the physiological determinants of the NA concentration measured. NA overflow from an organ gives a good reflection of nerve activity in that organ. However, sympathetic nerve activity is highly differentiated, particularly during stress, and conventional plasma NA levels (usually forearm venous samples) cannot be taken as an indication of 'sympathetic tone' in the whole individual. NA is rapidly removed from plasma, resulting in meaningless net veno-arterial concentration differences over organs unless its removal from arterial plasma is taken into account. In the forearm, for example, 40-50% of catecholamines are removed during one passage; about half of the NA in a venous sample is derived from the arm and half from the rest of the body. Therefore, conventional venous sampling overemphasizes local (mainly skeletal muscle) nerve activity. Whole-body sympathetic nerve activity may be monitored in arterial or mixed venous (i.e. pulmonary arterial) samples, which reflect NA overflow from all organs in the body. NA levels are determined both by overflow to plasma and clearance from plasma. NA turnover studies with 3H-NA infusions may be needed to assess clearance, but the simpler concentration measurements usually yield adequate information if the sampling site is relevant. NA overflow from an organ can be assessed (using 3H-NA or ADR as a marker for NA extraction in the organ) and provides valuable information on local sympathetic activity. Mental stress elicits marked circulatory responses, with mainly cardiorenal sympathetic activation and minor elevations of conventional venous plasma NA levels, thus illustrating the differentiated firing pattern of the sympathetic nerves. Circulating ADR is less important than neurogenic mechanisms in the responses to stress. Concentration-effect studies for infused catecholamines may be used for receptor sensitivity studies in vivo, but reflexogenic contributions to responses need to be determined. However, prejunctional mechanisms cannot be assessed without knowledge of the nerve activity present; for example, ADR infusion leads to increased nerve activity. When correctly sampled, measured and interpreted, plasma catecholamines can yield very valuable information on sympathoadrenal activity.