The validity of psychiatric diagnostic classifications is closely tied to their ability to specify etiology. However, due to limited knowledge of causal or susceptibility factors for most psychiatric disorders with a familial-genetic basis, current classification systems fail to achieve this ideal. This paper examines the difficulties in defining schizophrenia using ICD-10 and DSM-III-R criteria, highlighting discrepancies between diagnostic boundaries and observed familial aggregation patterns. Progress in understanding psychiatric disorders requires a foundation in epidemiological studies and an awareness of classification limitations. Clinical diagnoses aim to delineate distinct entities with specific etiologies. Since many psychiatric disorders have a familial-genetic basis, diagnostic definitions should reflect these pathways. When causal factors are known, definitions are directly derived from the phenotype, ensuring validity. However, detailed knowledge of these factors remains elusive for most disorders, except for Alzheimer's disease subtypes. Alternative strategies are needed to define disorders with familial-genetic origins. Two criteria for validity are proposed: higher genetic determination and stronger familial aggregation. These were used to establish the classification of affective disorders, distinguishing bipolar disorder from unipolar depression. Twin and family studies showed higher heritability for bipolar disorder and specific familial aggregation patterns. This led to the current DSM and ICD classifications. Bipolar II disorder, an intermediate syndrome between unipolar depression and bipolar disorder, was defined based on family studies showing specific aggregation patterns. However, current classification systems, including DSM-III-R, DSM-IV, and ICD-10, classified bipolar II under bipolar disorder. Familial-genetic studies have influenced the classification of schizoaffective disorders. The RDC first defined this disorder, showing it aggregates in families but not in a specific manner. Variants coaggregated with schizophrenia or affective disorders, leading to distinctions in DSM-III-R and ICD-10. Studies applying validity criteria to schizophrenia and psychotic disorders showed that negative symptoms had higher familial similarity and genetic load than positive symptoms. However, recent DSM and ICD definitions prioritize positive symptoms, despite evidence suggesting otherwise. The familial aggregation of schizophrenia is consistent across DSM-III-R, DSM-IV, and ICD-10, despite differences in diagnostic criteria. Linkage studies have not revealed specific relationships to diagnostic categories, highlighting the limited value of current diagnostic definitions for identifying genetic vulnerabilities. Future research must consider extended sample sizes and more specific genotype-phenotype relationships.The validity of psychiatric diagnostic classifications is closely tied to their ability to specify etiology. However, due to limited knowledge of causal or susceptibility factors for most psychiatric disorders with a familial-genetic basis, current classification systems fail to achieve this ideal. This paper examines the difficulties in defining schizophrenia using ICD-10 and DSM-III-R criteria, highlighting discrepancies between diagnostic boundaries and observed familial aggregation patterns. Progress in understanding psychiatric disorders requires a foundation in epidemiological studies and an awareness of classification limitations. Clinical diagnoses aim to delineate distinct entities with specific etiologies. Since many psychiatric disorders have a familial-genetic basis, diagnostic definitions should reflect these pathways. When causal factors are known, definitions are directly derived from the phenotype, ensuring validity. However, detailed knowledge of these factors remains elusive for most disorders, except for Alzheimer's disease subtypes. Alternative strategies are needed to define disorders with familial-genetic origins. Two criteria for validity are proposed: higher genetic determination and stronger familial aggregation. These were used to establish the classification of affective disorders, distinguishing bipolar disorder from unipolar depression. Twin and family studies showed higher heritability for bipolar disorder and specific familial aggregation patterns. This led to the current DSM and ICD classifications. Bipolar II disorder, an intermediate syndrome between unipolar depression and bipolar disorder, was defined based on family studies showing specific aggregation patterns. However, current classification systems, including DSM-III-R, DSM-IV, and ICD-10, classified bipolar II under bipolar disorder. Familial-genetic studies have influenced the classification of schizoaffective disorders. The RDC first defined this disorder, showing it aggregates in families but not in a specific manner. Variants coaggregated with schizophrenia or affective disorders, leading to distinctions in DSM-III-R and ICD-10. Studies applying validity criteria to schizophrenia and psychotic disorders showed that negative symptoms had higher familial similarity and genetic load than positive symptoms. However, recent DSM and ICD definitions prioritize positive symptoms, despite evidence suggesting otherwise. The familial aggregation of schizophrenia is consistent across DSM-III-R, DSM-IV, and ICD-10, despite differences in diagnostic criteria. Linkage studies have not revealed specific relationships to diagnostic categories, highlighting the limited value of current diagnostic definitions for identifying genetic vulnerabilities. Future research must consider extended sample sizes and more specific genotype-phenotype relationships.