The high prevalence and chronic evolution of schizophrenia are responsible for a major social cost. The adverse consequences of such psychiatric disorders for relatives have been studied since the early 1950s, when psychiatric institutions began discharging patients into the community. According to Treudley (1946) "burden on the family" refers to the consequences for those in close contact with a severely disturbed psychiatric patient. Grad and Sainsbury (1963) and Hoenig and Hamilton (1966) developed the first burden scales for caregivers of severely mentally ill patients, and a number of authors further developed instruments trying to distinguish between "objective" and "subjective" burden. Objective burden concerns the patient's symptoms, behaviour and socio-demographic characteristics, but also the changes in household routine, family or social relations, work, leisure time, physical health.... Subjective burden is the mental health and subjective distress among family members. While the first authors referred to those problems which are deemed to be related to, or caused by the patient, Platt et al. (1983) tried to distinguish between the occurrence of a problem, its alleged aetiology, and the perceived distress, when developing the SBAS questionnaire. These authors also proposed separate evaluations of behavioral disturbance and social performance by relatives, and a report of extra-disease stressors in family life. The SBAS is actually the most complete, but also complex instrument for evaluating burden in caregivers. Since 1967 Pasamanick and others proposed questionnaires for burden evaluation in relatives of schizophrenic patients. Relatives may be included in specific psychoeducational programs, but few of these programs have been evaluated in terms of caregiver burden. The theoretical approach in schizophrenia was not different from that one adopted in mentally ill population. Some instruments were validated first in a mentally ill group and then adapated for schizophrenic population. This paper describes the available data about intruments measuring caregiver burden in relatives of schizophrenic patients. Measures are summarized according to purpose, content and psychometric properties. Sixteen instruments have been collected from the litterature (1955-2001), and 2 instruments developed for relatives of mentally ill have also been taken into account. A group of 5 instruments focuses on the measurement of behavioural disturbance in persons with schizophrenia as perceived by their family members. Eleven scales include behavioural disturbance in a more complete decription of objective burden. Thirteen questionnaires also report the subjective distress in caregivers. One instrument has been developed in french language. Few of these instruments have been developed from a verbatim and really describe the caregiver's point of view. Most of them rely on experts point of view or on previously published studies. The content and domains explored by these instruments are variable. The psychometric properties are poorly documented for a number of them and no information is published about responsiveness. Some validated instruments are the Perceived Family Burden Scale (PFBS) the Involvement Evaluation Questionnaire (IEQ) and the Experience of Caregiving Inventory (ECI). In past studies, researchers more or less agreed about the dimensions that comprise the family burden. There was less agreement with regard to the definition of objective and subjective burden, and quite no agreement about the theoretical approach to the concept of burden. The evaluation of behavioural disturbance should now be excluded from the objective burden dimension. It is a specific domain, both objective and subjective, which can be described as a stressor in the stress-appraisal-coping model. A good approach of this domain can be found in the PFBS. It comprises 24 items and the principal components analysis produces 2 factors ("active" and "passive"), explaining 35% of the variance, with good consitency and acceptable test-retest reliability. The evaluation is both objective (presence or absence) and subjective (induced distress). The Behavior Disturbance Scale (BDS) may also be taken into account, although it is less validated. This scale derives from the SBAS, modified as a self-questionnaire, with both objective and subjective evaluations of all items. The concept of burden was recently modified in a new theoretical approach by Schene, when developing the IEQ. According to this author, the burden scale is supposed to exclude stressors (patient's behaviors), as well as outcome variables (distress or psychological impairment in caregiver). The "caregiving consequences" section comprises 36 items, which focus on the subjective aspects of the caregiver's experience. Principal component factor analysis generates 4 factors which explain 45% of the variance: tension, supervision, worrying, urging. The overall caregiving score substantially explains the connection between patient, caregiver, relationship variables and the caregiver's distress. This scale is a valid and simple instrument for caregiving eveluation The ECI also introduces a new approach of caregiving and rejects the notion of burden. The 66 item version is composed of 10 factors (8 "negative" and 2 "positive") with good internal consistency. The introduction of two positive factors (rewarding personal experiences, good aspects of the relationship with the patient) might be the basis of a useful outcome measure for intervention aimed at promoting caregiver well-being. Nevertheless, the authors fail to develop an overall score that includes these factors, and focus on the negative dimensions as predictors of morbidity and well-being. None of the variables included in the regression model explain a significant percent of the variance of the ECI positive score. None of these instruments was employed for evaluating programs or treatments, even psychoeducational programs for caregivers. This may be partly due to the lack of data about sensitivity to change. No instrument is now available for evaluating therapeutics from the caregiver's point of view. Developing such an instrument is necessary, in view of the increasing role of families in care for schizophrenic patients. These data and the review of the literature leeds us to propose the development of a self-administered questionnaire for evaluating subjective health-related quality of life in caregivers of schizophrenic patients. The instrument should be developed from the caregiver's point of view and be derived from qualitative interviews with relatives of patients suffering from schizophrenia. It's responsiveness should be documented before inclusion in clinical trials or evaluation of psychoeducational programs. We are now working with the National Union of Friends and Families of Patients to validate an instrument in french language.