The sweat test, even if carried out by an experienced technician, sometimes lacks reproducibility owing presumably to physiological variations (patient's diet, temperature, and other factors at present unrecognized). Some patients are particularly prone to exhibit this variability and in them a single sweat test is almost valueless. The aldosterone status is believed to be responsible for a reciprocal relationship between sweat sodium and potassium concentrations: tests done on 8 patients show that a high sweat potassium is associated with a correspondingly lower sodium--a circumstance which must be borne in mind when interpreting a patient's sweat sodium. Of 30 patients presenting with a variety of symptoms compatible with a diagnosis of cystic fibrosis and with sweat sodium ranging from 50 to 75 mEq/1 (50-75 mmol/1), only 4 have proved to have cystic fibrosis after several years of observations; 13 have later been diagnosed as having asthma. The problem of the 'grey area' of uncertainty is aggravated by the heterozygous state which is also associated with a sweat sodium in this range. Repeated sweat tests are indicated if the sweat sodium lies within the 'grey area', and the diagnostic importance accorded the test should diminish as the sodium value approaches this area. The diagnosis of cystic fibrosis must remain in doubt unless there is strong supportive clinical evidence.