For over half a century, pyridine-4-carboxy hydrazide (isonicotinyl hydrazide; isoniazid - INH) has been a front-line weapon in the battle against tuberculosis. Its metabolism has been the subject of important research, much of which has focused on the pharmacodynamic and toxicological aspects of certain INH metabolites. Since 1952, when the drug was first introduced, multiple INH metabolites have been identified, including hydrazine (HZ), isonicotinic acid (INA), ammonia, the acetylated derivative N(1)-acetyl-N(2)-isonicotinylhydrazide (AcINH), hydrazones with pyruvic and ketoglutaric acids (INH-PA and INH-KA, respectively), monoacetylhydrazine (AcHZ), diacetylhydrazine (DiAcHZ), and oxidizing free radicals. Their formation is the result of hydrolysis (INA, HZ), cytochrome P450 (CYP)-dependent oxidation (HZ, NH(3), oxidizing free radicals), and N-acetyltransferase (NAT) activity (AcINH, AcHZ, DiAcHZ). Doubts remain about isonicotinamide (INAAM) as an INH metabolite in mammals. Quantitatively speaking, one of the major metabolites is AcINH, which is produced by the enzyme NAT. It has virtually no antitubercular activity and is far less toxic than INH. Its formation and elimination are genetically controlled, and its elimination profile is trimodal (rapid, intermediate, and slow acetylation). Slow acetylation, which is transmitted as an autosomal recessive trait, increases the risk for peripheral neurotoxicity and hepatotoxicity in INH users. Thus far, there is no conclusive pharmacogenetic evidence that the formation of HZ and oxidizing radicals are linked to CYP polymorphisms. This article examines INH, HZ and its mono- and diacetylated metabolites, and ammonia (which in vitro and in vivo studies indicate as another derivative of HZ) in terms of their potential to cause neurotoxic and hepatotoxic effects (the two major forms of INH toxicity observed in animals and humans). INH hepatotoxicity seems to be related mainly to HZ, AcHZ, and other HZ metabolites that are capable of generating free radicals. The pathological aspects of slow INH acetylation will be discussed in relation to the drug's hepato- and neurotoxic effects. The mechanism underlying INH neurotoxicity has yet to be fully defined. The metabolite(s) involved in this phenomenon remain obscure although a major role is clearly played by HZ (and possibly also by the ammonia it releases). There is some evidence of the involvement of gamma-glutamyl HZ and of a chemical analogue of a Schiff base formed by INH and pyridoxal-phosphate. Recent findings have also revealed important interactions between INH and the various isoforms of CYP, and these may play a role in clinically relevant interactions between INH and several other drugs. All of these aspects of INH will be covered in the review.