The discrimination of borderline syndromes from the psychoses is often a difficult task clinically. The problem most often arises in the acute setting in which a crisis has arisen--the typical example being that of acute hospitalization. The clinician's task is to assess the patient's apparently psychotic symptoms and behaviors to determine whether they are the manifestations of an underlying psychotic process, or whether they reflect a more or less transient regression from a somewhat higher level of habitual functioning. Some discrimination between these categories is possible even in the acute presentation, since borderline patients only exceptionally demonstrate Schneiderian first-rank symptoms or any other discriminating indices of psychosis. While the differentiation may be clear cut between the psychotic and the higher-order, better functioning borderline, there may be less precision in discriminating between the lower-order borderline forms or transient borderline states and psychoses. We have focused on this area of differentiation in this study. The discriminating indices are both short- and long-term. The differentiation cannot be adequately made without longer-term evaluation of the patient. Nonetheless, on a short-term basis, evaluation of the patient's behavior can point the diagnosis in one direction or other. The presence of a clear precipitant; the presence of intense (often verbalized) anger; the patient's attempts to engage the therapist in an intense, dependent, clinging and demanding relationship, usually in manipulative fashion; the partial, fragmentary, often circumscribed and ego-alien quality of the patient's psychotic productions; the marked tendency to act-out feelings, particularly anger, in a way that gains increased attention and concern from doctors, family, friends, or hospital staff; the persistence of some degree of reality testing and areas of significant realistic functioning; the transient nature of regressive manifestations and the ready reversal of regression in structured environments and with appropriate therapeutic management, particularly adequate limit-setting--all point toward a borderline diagnosis. Moreover, these factors carry an accumulative weight so that the more of these factors that can be validated, the more secure the diagnosis of borderline psychopathology. On a longer-term basis, beyond a few days, one would expect the above indices to be better discriminated. In addition, there is greater opportunity to study patterns of patient behavior--both his interaction with staff and other patients and with the therapist.(ABSTRACT TRUNCATED AT 400 WORDS)