Recently, the National Association for Public Health Statistics and Information Systems considered changing the definition of gestational age from the current definition based on mother's last normal menstrual period (LMP) to the clinical/obstetric estimate determined by the physician (CE).They determined additional information was needed. This study provides additional insight into the comparability of the LMP and CE measures currently used on vital records among births at risk for poor outcomes. The data consisted of all New York State (NYS) (excluding New York City) singleton births in 2005 among mothers enrolled in the NYS Women Infants and Children (WIC) program during pregnancy. Prenatal WIC records were matched to NYS' Statewide Perinatal Data System. The analysis investigates differences between LMP and CE recorded gestations. Relative risks between risk factors and preterm birth were compared for LMP and CE. Exact agreement between gestation measures exists in 49.6% of births. Overall, 6.4% of records indicate discordance in full term/preterm classifications; CE is full term and LMP preterm in 4.9%, with the converse true for 1.5%. Associations between risk factor and preterm birth differed in magnitude based on gestational age measurement. Infants born to mothers with high risk indicators were more likely to have a CE of preterm and LMP full term. Changing the measure of gestational age to CE universally likely would result in overestimation of the importance of some risk factors for preterm birth. Potential overestimation of clinical outcomes associated with preterm birth may occur and should be studied.