All the available evidence from randomised controlled studies indicates that antipsychotic medications substantially reduce the risk of relapse in the stable phase of illness. The lowest dose should be chosen at which preferably no side effects occur, the risk of relapse seems to be optimally reduced and, if symptoms are still present, suppression of these is optimised. Side effects have to be assessed and, if necessary, pharmacotherapy has to be adjusted. Despite several methodological design issues, second-generation antipsychotics have proven similar or superior efficacy in preventing relapse and suppression (or even improvement) of symptoms compared to FGAs (available studies of the specific agents supply evidence for periods of up to 2 years). Due to the decreased risk of EPS, especially tardive dyskinesia, and, as observed in most studies, the superior efficacy in improving negative, cognitive and depressive symptoms together with at least comparable (for some agents, e.g., risperidone, olanzapine, superior) efficacy in relapse prevention, second-generation antipsychotics should be preferred in long-term treatment. Given all the known problems in compliance and discontinuation, which were underlined by the CATIE study, depot preparations should be considered for optimum effectiveness in preventing relapse. Randomised, control-group studies to determine the advantages of depot preparations of atypical neuroleptics compared to depots of typical neuroleptics are still lacking. The target strategy in long-term treatment of schizophrenia should be a combination of long-term antipsychotic treatment and psycho- and sociotherapeutic procedures, so that the relapse rate is further reduced and the course of disease can be further improved.