The ability of 21 drugs containing an amine function to form inhibitory P450-iron-metabolite complexes absorbing around 455 nm was studied on liver microsomes from rats treated with various P450 inducers. These drugs belong to different chemical and therapeutic series and exhibit very different structures. In the case of eight compounds (diltiazem, lidocaine, imipramine, SKF 525A, fluoxetine, L-alpha-acetylmethadol, methadol and desmethyltamoxifen) whose oxidation by microsomes from rats treated with several inducers was studied, only dexamethasone (DEX)-treated rat microsomes and, to a lesser extent, phenobarbital (PB)-treated rat microsomes, were able to give significant amounts of 455 nm absorbing complexes. Ten of the 21 compounds studied gave such complexes with DEX-treated rat microsomes, while only three compounds gave complexes (in low amounts) with PB-treated rat microsomes only. For all compounds leading to complexes both with DEX- and PB-treated rat microsomes, much higher amounts of complexes were obtained with DEX-treated rat microsomes. DEX-treated rat microsomes also led to the most intense type I spectral interactions with most of the compounds studied, and very often exhibited the highest N-dealkylation activities towards the tertiary or secondary amine function of the drugs used. A few exceptions aside, there generally exists a qualitative relationship between the ability of P450 3As, induced by DEX, to bind and N-dealkylate amino compounds and their propensity to lead to 455 nm absorbing complexes. This was confirmed by in vivo experiments showing that rats treated with diltiazem, tamoxifen or imipramine accumulated large amounts of 455 nm absorbing complexes in their liver only after pretreatment with DEX and, to a lesser extent, with PB. This particular ability of P450 3As to oxidize amino drugs with formation of inhibitory P450-metabolite complexes could be of great importance for the appearance of drug interactions in man.