Regression analysis tested the stated hypotheses using survey data collected from 275 family members of residents in 10 Dallas-Ft. Worth area nursing homes. This study is among the first to quantify family geographic distance, finding that family geographic distance is a significant negative predictor of nursing home contact.Additionally, results build on Allport's theory by extending its' usefulness to nursing home organizations in two distinct ways. First, findings support Allport's premise that contact alone between groups -i.e., family members and nursing home staff -is insufficient for increasing or decreasing family perceptions of nursing home care.Second, three of the four contact conditions included in Allport's theory were statistically supported by the data. In sum, findings of this research provide nursing homes with an empirically tested model for improving family perceptions of quality nursing home care.