The O-dealkylations of ethoxyresorufin and pentoxyresorufin are widely used activity probes for measuring the cytochrome P450 forms, CYP1A1 and CYP2B1, respectively, and their induction by xenobiotics, but there is confusion in the literature about which P450 forms are detected in human and rat liver microsomes by these and homologous alkoxyresorufins. High performance liquid chromatographic analysis confirmed that O-dealkylation to resorufin was the sole or predominant route of metabolism for both short-chain and long-chain alkoxyresorufins and benzyloxyresorufin by rat liver microsomes. The purified 3-methylcholanthrene (3MC)-induced rat P450 forms, CYP1A1 and CYP1A2, and a possible variant form, CYP1A1*, showed substrate selectivities for propoxyresorufin, methoxyresorufin and ethoxyresorufin, respectively. Purified phenobarbitone (PB)-induced CYP2B1 was selective for benzyloxyresorufin and pentoxyresorufin. Purified constitutive CYP2C6 was much less active than CYP2B1 or the CYP1A forms but showed distinctive selectivity for benzyloxy-, propoxy- and butoxyresorufin. CYP1A2 and CYP2C6 metabolised n-propoxy- and n-butoxyresorufin much more rapidly (8-23-fold) than iso-propoxy- and iso-butoxyresorufin, whereas CYP1A1 and CYP2B1 showed only small differences (2-5-fold) between the n- and iso-homologues and CYP1A1* and CYP2B2 did not discriminate between them. The results show that ratios between different alkoxyresorufin O-dealkylation (AROD) activities can be more useful than absolute values of single activities for identifying P450 forms. Anti-P450 antibody and furafylline inhibition of rat liver microsomal AROD confirmed that ethoxyresorufin was a selective probe for CYP1A1 in 3MC-induced and isosafrole (ISF)-induced microsomes and that pentoxy- and benzyloxyresorufins both selectively measured CYP2B1 in PB-induced and ISF-induced microsomes. Ethoxyresorufin was not a selective probe for CYP1A in liver microsomes from untreated or PB-induced rats, however, where it was metabolised mainly by CYP2C6 and CYP2B1, respectively. Pentoxyresorufin and benzyloxyresorufin were metabolised by several different P450 forms in non-induced rat liver microsomes but mainly by the CYP1A subfamily in 3MC-induced microsomes and by CYP2B1 in PB- and ISF-induced microsomes. Although with purified rat P450s methoxyresorufin appeared not effectively to discriminate CYP1A2 from CYP1A1, CYP1A1* or CYP2C6, furafylline inhibition indicated that methoxyresorufin was a selective measure of CYP1A2 in uninduced and 3MC-induced rat liver microsomes but not in ISF- or PB-induced microsomes. In human liver microsomes, antibody inhibition and furafylline inhibition showed that ethoxyresorufin and methoxyresorufin were metabolised mainly by CYP1A2, whilst benzyloxyresorufin metabolism was due mainly to the CYP3A subfamily but also involved CYP1A2 and CYP2A6. There was considerable interindividual variation in the roles of different P450 forms in all three reactions in human liver.