Grade 3 intraventricular hemorrhage (IVH) (without parenchymal involvement) and grade 4 IVH (with parenchymal involvement) are often combined into description of a single entity, usually "severe" IVH, despite different long-term neurodevelopmental outcome. Although risk factors for severe IVH have already been well described, it is not known if these risk factors and associated short-term neonatal morbidities are different for grade 3 and grade 4 IVH, and indeed, this clustering of grade 3 and grade 4 IVH into severe IVH precludes further delineation of the potential risk and protective factors that can be altered to reduce the incidence of grade 4 IVH, which is presumably associated with worse outcome compared with grade 3 IVH. We sought to characterize and compare commonly cited risk factors and associated short-term neonatal morbidities between grade 3 and grade 4 IVH in very low-birth-weight (VLBW) infants. We performed a retrospective review of VLBW (birth weight < 1500 g) infants with severe IVH born between January 2001 and March 2007. Fifty-nine (10.5%) of 562 infants surviving beyond 3 days of age had severe IVH as recorded on routine cranial sonography during the first 7 to 10 days of life, 28 had grade 3, and 31 had grade 4 IVH. Infants with grade 4 IVH were younger [gestational age (weeks), grade 4 IVH versus grade 3 IVH: 25.5 +/- 1.7 versus 26.7 +/- 1.7, p = 0.02) and weighed less at birth [birth weight (g), grade 4 IVH versus grade 3 IVH: 860 +/- 214 versus 1007 +/- 253, p = 0.03) compared with infants with grade 3 IVH. Other commonly cited clinical factors that alter the risk for severe IVH, including mode of delivery, pregnancy-induced hypertension, premature and/or prolonged rupture of membranes, maternal fever, maternal bleeding, prenatal steroid administration, maternal magnesium sulfate therapy, 1-minute and 5-minute Apgar scores, need for delivery room resuscitation (epinephrine and chest compressions), surfactant therapy, presence of refractory hypotension, evidence of early onset culture-proven sepsis, use of high-frequency ventilation, presence of pneumothorax, and hemodynamically significant patent ductus arteriosus, were similar between infants with grade 3 and grade 4 IVH. Carbon dioxide tensions (minimum PaC (2), maximum PaCO(2), mean PaCO(2), standard deviation of PaCO(2), and coefficient of variation of PaCO (2)) in infants receiving mechanical ventilation during first 3 postnatal days were also not statistically dissimilar. To determine the variables differentiating grade 3 from grade 4 IVH in the study population, logistic regression analysis confirmed only the independent association of gestational age (odds ratio [OR] 0.6, 95% confidence interval [CI] 0.5 to 0.9, P = 0.012) and maternal magnesium sulfate therapy (OR 0.3, 95% CI 0.07 to 0.9, P = 0.04) with the development of grade 4 IVH. Short-term neonatal morbidities were also similar between infants with grade 3 and grade 4 IVH. Among VLBW infants, the risk of a grade 4 versus grade 3 IVH increases with declining gestational age, but does not appear to be related to other commonly cited clinical factors. This information may be useful for prognostication and may improve the quality of parental counseling.