HOP, defined as normal TSH and low T4 in preterm infants, has been associated with poor longterm neurodevelopmental outcome, implying that treatment with thyroxine might be beneficial. Conversely, lower metabolic rates induced by low T4 may be protective. Chart review of 893 infants≤1500 gms born 1/1/93-12/31/95 revealed 316 with HOP by state screen, with the remaining 577 serving as controls. Infants with HOP were younger(27.2±2.4 vs 28.3±3.1 wks EGA), smaller (BW 960±264 vs 1067±283 gms), and morbidity greater for the following markers: Gd 4 IVH, 7 vs 4.5%; Gd 3 IVH, 9.5 vs 4.8%; NEC, 20.6 vs 9.6%; NEC + surgery, 15.2 vs 4.5%; late sepsis, 45.6 vs 26.9%; PDA ligation, 27.2 vs 8.8%; vent-dependency: 8-30 days, 39.2 vs 18.8%; 31-90 days, 20.3 vs 4.1%; >90 days, 2.8 vs 0.8% (each p<0.001, except Gd 4 IVH, p=0.08). Mortality was not dif- ferent between HOP and controls (17.7 vs 19.2%; p=0.59). However, if controlled for EGA, BW and gender, mortality was lower in HOP infants (odds ratio 55%, p=0.003). These data indicate that HOP occurs in smaller and sicker infants, suggest that HOP may protect against mortality, and emphasize the need for controlled trials examining the effects of thyroxine treatment on shortterm mortality as well as longterm neurodevelopmental outcome.