Objective. This study examined to what degree patient-centeredness—measured as an underlying ability of obstetrical nurses—influenced Medicaid patients’ satisfaction with care in hospital obstetrical units. Design. Multigroup structural equation modeling design, using three cross-sectional random samples (n=300 each) from the 2003 Press Ganey National Inpatient Database. Setting. Self-administered mail surveys. Participants. 900 Medicaid recipients recently discharged from inpatient hospital obstetrical units across the United States. Methods. Multigroup structural equation modeling was used to test the goodness of fit between a hypothesized model based on the Primary Provider Theory and patients’ ratings of nurses. Results. The model fitted the data well, was stable across three random samples, and was sustained when compared to a competing model. The patient-centeredness of nurses significantly influenced overall patient satisfaction and explained 66% of its variability. When nurses’ patient-centeredness increased by one standard deviation, patients’ satisfaction increased by 0.80 standard deviation. Conclusion. This study offers a novel approach to the measurement of the patient-centeredness of nurses and a paradigm for increasing it and its influence on Medicaid patients’ satisfaction in hospital obstetrical units.