Discrete, bilateral, radiofrequency destruction of the supra-optic nucleus resulted in a parallel fall in levels of immunoreactive (ir) dynorphin (DYN), ir-α-neoendorphin (α-NE) and ir-vasopressin (VP) in the hypothalamus, neuroin-termediate lobe (NIL) of the pituitary and septum of rats, whereas in the medulla/pons, midbrain and anterior lobe (AL), levels of these peptides were not significantly changed. The content of ir-β-endorphin (β-EP) was, in contrast, elevated in the hypothalamus and unchanged in the NIL and AL, while that of ir-metenkephalin (ME) was modified in neither the hypothalamus nor medulla/pons. As evaluated in lesioned and sham rats, collectively, hypothalamic levels of ir-DYN, ir-α-NE and ir-VP were positively correlated with their counterparts in the NIL but not, with the exception of ir-VP, with those in the AL. Ir-β-E did not show these relationships. Within the NIL and in the hypothalamus, levels of ir-DYN, ir-α-NE and ir-VP were positively correlated with each other and independent of those of ir-β-EP and ir-ME. In the AL and other brain tissues, however, only those of ir-DYN and ir-α-NE were positively and significantly correlated in contrast to ir-VP. Further, across all structures, although the ratio of levels of ir-α-NE to ir-DYN was quite constant, that of ir-VP to ir-DYN varied enormously. These data indicate that:(1) the supra-optic nucleus may contribute to NIL, hypothalamic and septal but not to AL, midbrain or medulla/pons pools of ir-VP, ir-DYN and ir-α-NE, and that it interacts with those of ir-β-EP in the hypothalamus. (2) in the NIL and hypothalamus, ir-DYN, ir-α-NE and ir-VP are, as compared to ir-β-EP and ir-ME, closely interrelated. (3) AL pools of DYN and ir-α-NE are closely coupled but, in contrast to NIL pools, not related to ir-VP and independent of the hypothalamus. (4) Extrahypothalamic brain pools of ir-DYN and ir-α-NE may be not related to ir-VP. The possible existence of ir-DYN and ir-α-NE independently or ir-VP in the AL and extrahypothalamic nervous tissue is discussed.