Extracts of mammalian atrial tissue contain potent natriuretic substances known collectively as atrial natriuretic factor (ANF). The purposes of the present experiments were: 1) to improve on existing bioassay methodology for the detection of ANF activity in atrial extracts, and 2) to compare the ANF activity of atrial extracts prepared from Brattleboro-stain diabetes insipidus (DI) rats with that from normal and water-deprived Long-Evans (LE) rats. A pool of atrial tissue extract (AE) was prepared from normal Sprague-Dawley rats for use as a standard against which unknown AE samples could be compared. Five doses, ranging from 27 to 432 micrograms of AE protein, were assayed in the Sprague-Dawley bioassay rats. Phosphate-buffered saline (PBS) vehicle and ventricular tissue extracts were also assayed. Statistical analysis of several log dose-response relationships revealed that the bioassay response most appropriate in determining relative natriuretic activity of AE was the log of the experimental/control ratio for sodium excretion. The bioassay was used to demonstrate that PBS atrial extracts from both water-deprived LE rats and DI rats contain more natriuretic activity than do PBS atrial extracts from LE rats.