We evaluated the antecedents of CP in VLWB infants (600-1250 g) enrolled in a multicenter trial of low-dose indomethacin for the prevention of intraventricular hemorrhage (IVH) in order to identify potential CP prevention strategies. Neurologic examinations were performed on 417 of the 440 survivors(95%) at 18 to 36 mos. corrected age looking for classical forms of CP. Multiple pre-, peri- and postnatal variables were recorded including the results of frequent cranial ultrasound scans obtained throughout the newborn period and interpreted by at least 2 readers. Multivariate analyses were performed. CP was found in 35 of 417 subjects (8.4%); the types were equally divided between diplegia (N=11), hemiplegia (N=12) and tetraplegia (N=12). Grades 3-4 IVH, posthemorrhagic hydrocephalus, BPD, periventricular leukomalacia (PVL) and ventriculomegaly (VM) on the 40 week post-conception age ECHO were all univariate predictors of CP (p < 0.001 for all). PVL and VM were associated with the highest detection rates for CP (35% and 40%, respectively) with low false positive rates (4% and 6%, respectively).Multivariate step down analysis of pre-, peri- and postnatal variables showed male sex (p = 0.01), five minute Apgar score (p = 0.001), PVL(p = 0.002), & VM (p = 0.0001) were independent & important predictors of CP. Analysis for effects of antenatal MgSO4 exposure was NS [9/141(6%) exposed vs 26/276 (9%) unexposed,p = 0.35]. We conclude the CP rate is low in VLBW infants and IVH and its complications represent major predictors of CP in this population. (supp by NS 27116)