Background: Over the last century, especially during the latter half, the prevalence of the diagnosis of catatonic schizophrenia decreased considerably. Several explanations for this phenomenon have been put forward. Sampling and Methods: The present study investigated the frequency of the diagnosis of catatonic schizophrenia in a large sample of admitted psychiatric patients (n = 19,309). In addition, the presence of catatonic symptoms was studied in a sample of patients with schizophrenia (n = 701) and in a group of consecutively admitted psychotic patients (n = 139). In these two groups the effect of the diagnostic procedures on the recognition of catatonia was examined. Results: The diagnosis of catatonic schizophrenia dropped from 7.8% in 1980–1989 to 1.3% in 1990–2001 (p < 0.001). In addition, a possible under-diagnosis of catatonic schizophrenia was found in an independent sample of patients with schizophrenia. Application of a systematic catatonia rating scale in patients admitted with acute psychosis identified a bimodally distributed catatonic dimension. At least 18% of these patients fulfilled the criteria for catatonia. Interestingly, the catatonic subgroup used atypical antipsychotic compounds more frequently (p < 0.05). Conclusions: The results suggest that changes in diagnostic criteria and the diagnostic procedure itself are responsible for the under-recognition of catatonia.