The recurring digit span test is hypothesized to assess consolidation memory and hypothetically addresses neuropsychological left-temporohippocampal functioning, an area of the brain which is presumed to play a crucial role in schizophrenia. The present study applied the recurring digit span memory test and the recurring block span memory test (assumed to be an indicator of right-temporohippocampal functioning) to 106 acute schizophrenic patients (DSM-III-R), 49 siblings of these patients without DSM-III-R lifetime diagnoses as well as 53 non-related control subjects without DSM-III-R lifetime diagnoses. Results of our investigation indicate that schizophrenic patients differ from control subjects regarding recurring digit span in a reduced number of correct reproductions of recurring trials. Since performance of schizophrenic patients was only non-significantly different from control subjects regarding recurring block span memory, our findings point to a task-specific memory impairment in schizophrenia and argue against a generalized deficit in this disease. Furthermore, individuals at risk for schizophrenia also differed from control subjects concerning the recurring digit span memory test but not with regard to the recurring block span test. Therefore we conclude that the reduced cumulative learning curve of verbal material (reduced number of correct reproduction of recurring digit trials) in schizophrenic patients and their healthy siblings might represent a marker for the liability to schizophrenia and parallels neuroimaging findings concerning left-sided temporohippocampal abnormalities in this disease, which presumably play a key role in the etiology of schizophrenia.