As part of an epidemiological study of congenital hydrocephalus in Utah, we focused on the effect of ascertainment sources and temporal variability to further delineate the causes of this relatively common, handicapping birth defect. The incidence and distribution of 934 reported cases diagnosed prior to age 6 months, and born to Utah residents from 1940 to 1979, were analyzed. Data were ascertained by examination of multiple sources, e.g., 982,066 birth, 11,161 fetal death, and 248,208 death certificates, and selected hospital and clinic records. Of the 934 reported cases, 700 met our selection criteria for congenital hydrocephalus, which results in a crude incidence of 0.70 per 1,000 live and stillbirths. Seventy-one cases (10.1%) had additional, multiple congenital anomalies. The male/female sex ratios of the 619 cases of isolated congenital hydrocephalus (occurring as a single entity or in the absence of other reported or known birth defects) and those with multiple congenital anomalies (71 cases) were virtually identical, being 1.45 and 1.48, respectively. A significant 85% increase in the rate of reported cases was observed for the period 1966 to 1970. However, examination of patients' records from 1966 to 1975 in the hospital responsible for almost all of this increase suggests that this was an iatrogenically related "epidemic" caused by several factors: the introduction and possible misinterpretation of pneumoencephalograms (PEG) in the diagnosis of hydrocephalus (PEG was replaced by CAT scanning in the early 1970s), inappropriate diagnosis, and incorrect recording of age at time of diagnosis.