BACKGROUND: An increasing interest in the study of cognition in Schizophrenia has developed within the last few years although cognitive problems have been described in this disorder since the beginning of the 20th century. Presently, various data tend to assert that cognitive disorders are the core disturbance in schizophrenia and that their severity is predictive of the course of the disease. Indeed, studies have shown that the disturbances measured in cognitive tests are neither the consequences of positive or negative symptoms, nor related to motivation or global intellectual deficit, nor to anti-psychotic medication. It is also presently known that the severity of cognitive symptoms is a better indicator of social and functional outcome than the severity of the negative or positive symptoms. The patients who have the most severe cognitive deficits during the first episode of the disease are most likely to present a chronic and severe form later on. The aspects of cognition that are specifically impaired in schizophrenia are verbal memory, working memory, motor function, attention, executive functions, and verbal fluency. Cognitive disturbances are thus very important in several fields of research in schizophrenia such as: understanding the psychopathology, epidemiology (indicators of vulnerability), genetics (endophenotypes), neuro-imaging (including functional neuro-imaging), and psychopharmacology (they can be used as a parameter of evaluation in therapeutic trials with new molecules, or cognitive psychotherapy). LIMITS OF COGNITION ASSESSMENTS: However, there are some methodological limits to these cognitive evaluations. First, schizophrenia is a heterogeneous disease and there are no specificities of the different subgroups in terms of cognition. Secondly, the time chosen to evaluate the abilities of the patient is also a limiting factor. But most of all, the batteries of tests used in different studies are not standardized. BRIEF ASSESSMENT OF COGNITION IN SCHIZOPHRENIA: It is therefore of great interest to create an available and easily used battery of validated tests. This would enable one to measure the different cognitive deficits and to repeat the tests, and assess evolution through longitudinal follow up of the patients. The BACS is a new instrument developed by Keefe et al. in the Department of Psychiatry and Behavioural Sciences at the University of Duke Medical Centre. It evaluates the cognitive dimensions specifically altered in schizophrenia and correlated with the evolution of the disease. This test is simple to use, requiring only paper, pencils and a stopwatch. It can be administered by different carers. The duration of the test session is approximately 35min. This battery of tests was validated on a sample of 150 patients compared with a sample of 50 controls, matched for age, parent education and ethnic groups. This aim of this study is to create a French adaptation of the BACS (translation and back translation approved by the Department of Psychiatry and Behavioural Sciences at the University of Duke Medical Centre) and then to test its easiness of administration and its sensitivity, performing correlation analysis between the French Version of the BACS (version A) and a standard battery. Its adaptation and validation in French would at first be useful for the French-speaking areas and then would add some new data for the pertinence of using the BACS. METHODS: 35 French stabilized schizophrenic patients were recruited from the inpatient and outpatient facilities at the Clermont-de-L'Oise Mental Health Hospital (Picardie area, France) in Dr Boitard's Psychiatric Department (FJ 5.) Patients were required to meet DSM-IV criteria for schizophrenia or schizoaffective illness. The patients were tested on two separate days by two independent clinicians with less than two weeks between the two assessments. During the first test session, subjects received the French A version of the BACS and during the second session, they were administered the standard battery of cognitive tests including: the Rey Auditory-Verbal learning test, the Wechsler Adult Intelligence Scale, third edition, subtests (Digit inverse sequencing, Digit Symbol-Coding), the Trail-Making A, Verbal Fluency (Controlled Oral Word Association Test, Category Instances), and the Wisconsin Card Sort Test (128 card version). The factor structure of the French BACS A Version was determined by performing a principal components analysis with oblique rotation. The relationship between the French BACS sub-scores and the standard battery sub-scores was determined by calculating Pearson's correlations among the sub-scores, with a level of significance of alpha<0.05. RESULTS: All the 35 patients completed the standard battery and each subtest of the French BACS A Version without interruption and with good understanding of the instructions. The average duration of the BACS test sessions was 36.51min (S.D.=12.14.) compared to the standard battery in which the sessions lasted more than one hour with more difficulty during the Wisconsin tests. The factor analysis conducted on the data collected from patients suggests that there is a single dimension, a factor of general cognitive performance, which accounted for the greatest amount of variance. The BACS thus permits an assessment of overall cognitive function as a global score, more than some individual specific cognitive domains. The sub-scores from the French BACS A Version were strongly correlated with the standard battery corresponding sub-scores. We observed significant correlations for all the subtests evaluating: verbal memory (Pearson=0.83; p<0.001; IC [0.69; 0.91]), working memory (Pearson=0.67; p<0.001; IC[0.43; 0.80]), verbal fluency (semantic: Pearson=0.64; p<0.001; IC[0.40; 0.80]), alphabetical (Pearson=0.87; p<0.001;IC[0.77; 0.93]), attention and speed of information processing (Pearson=0.69; p<0.001; IC[0.47; 0.83]), executive function (Pearson=0.64; p<0.001; IC[0.39; 0.80]). We almost found a significant correlation for motor speed (Pearson=-0. 32; p=0.06; IC [-0.59; -0.014]). CONCLUSION: The French adaptation of the BACS scale is easier to use in schizophrenic patients with French as mother tongue, with a completion rate equal to 1, and also with less than 35min to complete and check. We obtained significant correlations for all domains except motor speed, which is almost significant. The BACS is as sensitive to cognitive impairment in patients with schizophrenia as a standard battery of tests that required over 2h to complete. Moreover, these results demonstrate that the BACS, the global score of which may be the most powerful indicator of functional outcome, can also be a good neuropsychological instrument for assessing global cognition in patients with schizophrenia.