The clinical pharmacokinetics of the 4 depot antipsychotics for which plasma level studies are available (i.e. fluphenazine enanthate and decanoate, haloperidol decanoate, clopenthixal decanoate and flupenthixol decanoate) are reviewed. The proper study of these agents has been handicapped until recently by the necessity of accurately measuring subnanomolar concentrations in plasma. Their kinetic properties, the relationship of plasma concentrations to clinical effects, and conversion from oral to injectable therapy are discussed. The depot antipsychotics are synthesised by esterification of the active drug to a long chain fatty acid and the resultant compound is then dissolved in a vegetable oil. The absorption rate constant is slower than the elimination rate constant and therefore, the depot antipsychotics exhibit ‘flip-flop’ kinetics where the time to steady-state is a function of the absorption rate, and the concentration at steady-state is a function of the elimination rate. Fluphenazine is available as both an enanthate and decanoate ester (both dissolved in sesame oil), although the decanoate is more commonly used clinically. The enanthate produces peak plasma concentrations on days 2 to 3 and declines with an apparent elimination half-life (i.e. the half-time of the apparent first-order decline of plasma concentrations) of 3.5 to 4 days after a single injection. The decanoate produces an early high peak which occurs during the first day and then declines with an apparent half-life rangingfrom 6.8 to 9.6 days following a single injection. After multiple injections of fluphenazine decanoate, however, the mean apparent half-life increases to 14.3 days, and the time to reach steady-state is 4 to 6 weeks. Withdrawal studies with fluphenazine decanoate suggest that relapsing patients have a more rapid plasma concentration decline than non-relapsing patients, and that the plasma concentrations do not decline smoothly but may exhibit ‘lumps’ due to residual release from previous injection sites or multicompartment redistribution. Cigarette smoking has been found to be associated with a 2.33-fold increase in the clearance of fluphenazine decanoate. In 3 different studies, fluphenazine has been proposed to have a therapeutic range from < 0.15 to 0.5 ng/ml with an upper therapeutic range of 4.0 ng/ml. Plasma concentrations following the decanoate injection are generally lower than, but clinically equivalent to, those attained with the oral form of the drug. Haloperidol decanoate plasma concentrations peak on the seventh day following injection although, in some patients, this peak may occur on the first day. The apparent elimination half-life after multiple injections is approximately 3 weeks and the time to reach steady-state is approximately 3 months. The reduced metabolite of haloperidol is present in significant quantities in humans following oral therapy; however, this has not been reported in pharmacokinetic studies with the depot preparation. The therapeutic range for haloperidol has been reported to be 3 to 40 ng/ml in several studies, but these studies did not measure the reduced metabolite. It has been suggested that the presence of high concentrations of the reduced metabolite may affect the response to haloperidol. Clopenthixol is only active as the cis (Z) isomer, which is the form of the drug present in the decanoate preparation. Peak plasma concentrations are usually reached between 4 and 7 days after injection. Following multiple injections, the apparent elimination half-life is 19 days. No correlation between plasma concentrations and clinical response has been reported, but concentrations have been observed to range from 10 to 100 ng/ml following a wide range of dosages and injection intervals. Flupenthixol is available as either the palmitate or decanoate ester, although most pharmacokinetic studies have used the decanoate formulation. Like clopenthixol, flupenthixol is only active as the cis (Z) isomer which is the form present in the decanoate preparation. Peak plasma concentrations occur on approximately the seventh day following injection, and no pharmacokinetic differences have been observed between the 2% and 10% concentrations of the drug. Following a single injection, an apparent elimination half-life of 8 days was reported, whereas after multiple injections, the apparent half-life was 17 days. Following withdrawal, no statistically significant differences were noted between relapsing and non-relapsing patients, but there was a trend towards lower trough concentrations in the relapsing group. There was also a shorter apparent half-life in the relapsing group when compared with the non-relapsing patients. No correlations between plasma concentrations and clinical response were found. Most methods for converting oral preparations to depot injections have been empirically developed. One study that did examine plasma concentrations and clinical effects during conversion of oral fluphenazine to the decanoate found that 1.2 to 2.5 times the oral dose (mg/day) given weekly resulted in a smooth transition between dosage forms. Fluphenazine plasma concentrations on the decanoate were lower during the first 3 weeks in comparison with oral fluphenazine therapy. It is recommended that the initially effective dose be reduced or the injection interval increased after 4 to 6 weeks to prevent possible accumulation of drugs as plasma concentrations approach steady-state. The recommended dose of haloperidol decanoate is 20 times that of the daily oral dose (mg/day) given monthly, but this should be reduced to a factor of 15 in geriatric patients. Flupenthixol 10mg given orally daily is proposed to be equivalent to 25mg of the decanoate given weekly. The use of the mean conversion ratios as a starting point for an individual patient’s conversion is appropriate, but wide interindividual variations in pharmacokinetics require plasma level monitoring and careful clinical observation of the patient. Adverse effects from depot antipsychotics are relatively rare, except for extrapyramidal system reactions. The most serious (albeit very rare) problem that can occur with a long acting antipsychotic is the neuroleptic malignant syndrome. Due to their long apparent half-lives after injection, effective treatment of this syndrome following administration of depot antipsychotics can be difficult.