Pervasive developmental disorders (PDDs) are severe psychiatric disorders characterised by impaired social interaction, reduced verbal and nonverbal communication, and restricted and stereotyped patterns of interest and behaviour, with onset in the first 3 years of life. As part of a multimodal approach that includes educational and behavioural interventions, appropriate use of pharmacotherapy in patients with PDDs can improve some symptoms and behaviours and increase the patient's response to nonpharmacological interventions. Pharmacotherapy is aimed at stabilising the dysregulated systems that are thought to underlie the abnormal behaviours, namely the serotonergic and dopaminergic systems.This paper critically reviews the available literature on the pharmacotherapy of PDDs in children and adolescents. Pharmacological, efficacy and safety data on drugs acting on serotonin (serotonergic agents such as clomipramine, fluoxetine, paroxetine, sertraline, citalopram and fluvoxamine), on both serotonin and norepinephrine (venlafaxine and mirtazapine), dopamine (dopamine-blocking antipsychotics such as haloperidol and pimozide), and on both serotonin and dopamine (atypical antipsychotics such as clozapine, risperidone, olanzapine, quetiapine and ziprasidone) are presented. The paper concludes with a brief discussion of other psychotropic medications, such as the opioid antagonists, alpha-adrenoceptor agonists (clonidine and guanfacine), psychostimulants, mood stabilisers, glutamatergic compounds (lamotrigine, amantadine), buspirone, and secretin. Available evidence suggests that atypical antipsychotics (mainly risperidone and olanzapine) are particularly indicated when more serious behavioural symptoms, such as aggression, self-injurious behaviours and hyperactivity, are prominent. Serotonergic agents may be effective in children with repetitive phenomena or affective symptoms. Social relatedness is more frequently refractory to both atypical antipsychotics and serotonergic agents. Further research into the drug treatment of PDDs, including placebo-controlled trials and long-term, naturalistic follow-up studies of large samples of patients of different age ranges, is warranted.