A concept of an endophenotype, also termed as an internal endophenotype, is used in genetic studies on psychiatric disorders. Neurological soft signs are also considered candidates for endophenotypes of schizophrenia. Neurological soft signs are, objectively measured, non-localizing abnormalities, not related to impairment of a specific brain region, reflecting improper corical-subcorical and intercortical connections. This paper presents the main domains of NSS, methods of measurement of NSS, their neuroanatomical substrate, association of NSS with schizophrenia symptoms the and analysis of the literature in order to check whether NSS meet the criteria of the phenotype. A marker can be considered a phenotype if it meets the following criteria: 1) association with a disease in a population, 2) heritability, 3) state-independence, 4) familial association (the endophenotype is more prevalent in the affected individuals, their affected and non-affected family members in comparison to the normal population), 5) co-segregation (the endophenotype is more prevalent among ill family members of ill probands compared with healthy relatives). Currently, there is an ample evidence that the NSS, especially these representing impaired motor coordination, meet certain criteria of an endophenotype. However, there are still several unresolved questions concerning NSS: studies on relatives of schizophrenic patients included small groups of subjects, many of the studies included individuals with schizophrenia, as well as schizophrenia spectrum disorders, the available date-base of twins (schizophrenia-concordant and schizophrenia non -concordant) is not sufficiently large, there are too few studies evaluating the relationship of NSS and individual genes, there are no objective and quantitative methods of measurement of NSS. Therefore, NSS still represent only candidates for an endophenotype of schizophrenia. Finding correlations of selected NSS with other endophenotypes and their genetic correlates also needs further investigation and may provide a definitive answer to the question of the usefulness of NSS as the endophenotype of schizophrenia.