A critical processing step in endothelin biosynthesis is the conversion of the intermediate "big endothelin" to its biologically active product catalysed by endothelin converting enzyme (ECE). In this commentary we discuss critically the cellular location, structure, and activity of the isoforms of ECE. The current evidence supporting a metallopeptidase ECE as the physiological regulator of endothelin production is described. Its sensitivity to inhibition by the fungal metabolite phosphoramidon and subsequent cloning of the enzyme indicate it to be a type II integral membrane protein homologous with neural endopeptidase-24.11 (E-24.11), the major neuropeptide-degrading ectoenzyme in brain and other tissues. Unlike E-24.11, however, ECE exists as a disulphide-linked dimer of subunit M(r) 120-130 kDa and is not inhibited by other E-24.11 inhibitors such as thiorphan. Alternative splicing produces two forms of ECE with distinct N-terminal tails. These isoforms of ECE-1 show similar specificity converting big endothelin-1 (ET-1) to ET-1 but big ET-2 and big ET-3 are converted much less efficiently. This suggests that additional forms of ECE remain to be isolated. Immunocytochemical studies indicate a predominant cell-surface location for ECE-1, like E-24.11. This is consistent with the conversion of exogenous big ET-1 when administered in vivo and the inhibition of this event by phosphoramidon. However, mature ET-1 can be detected in intracellular vesicles in endothelial cells, suggesting that some processing occurs in the constitutive secretory pathway. This may be mediated by ECE-2, a recently cloned member of the E-24.11/ECE family which has an acidic pH optimum. Selective inhibitors of ECE may have therapeutic applications in cardiovascular and renal medicine.