Fifty-one sequential intubated babies with birth weights of less than 1,751 were evaluated by serial Doppler ultrasound during the first three days of life. These babies were part of a phenobarbital prophylaxis trial cohort study. Subependymal-intraventricular hemorrhage developed in 17 of the babies. Infants with subependymal-intraventricular hemorrhage, whether or not they received pancuronium or phenobarbital, had coefficients of variation comparable to those of babies without hemorrhage. Coefficient of variation values of the right were comparable to values obtained from the left anterior cerebral artery complex and did not appear to be consistently altered by the presence of subependymal-intraventricular hemorrhage. Coefficient of variation values appeared to be consistently greatest on day 1 and lowest on day 2. In addition, the values overall increased as the number of waves used to determine the coefficient of variation enlarged from five to 20. This phenomena, however, was not seen among pancuronium recipients and suggests that movement artifact may be a determinant of coefficient of variation values. We conclude that, when the best 20 waves are chosen to evaluate the coefficient of variation, no association exists between coefficient of variation values and development of subependymal-intraventricular hemorrhage or administration of phenobarbital.