The purposes of this study were first to determine if there were significant differences in living environment press, i.e., the pressure on an individual to behave in a certain way, between those fraternity chapters that had the highest freshman pledge retention rates and those that had the lowest. If significant differences were determined to exist, the second purpose was to investigate how those differences related to differences in the retention rates for freshman fraternity pledges. The data were obtained from the records of all freshman pledges in the Oregon State University fraternity system for a four-year period. From these data the high pledge retention and low pledge retention fraternities were determined. The sample for the remainder of this study was two of the three highest and two of the three lowest pledge retention fraternities. The highest and lowest pledge retention fraternities were eliminated. Analyses included: Pearson Correlation Coefficients to determine if there were significant correlations between retention of pledges in the fraternity system and six factors involving grades and the number of members and pledges living in the fraternities; two-way, fixed analyses of variance to determine if there were significant differences between the high retention fraternities (HRFs) and the low retention fraternities (LRFs) with respect to pledge high school grade point averages and Scholastic Aptitude Test scores; chi square contingencies to determine if there were significant differences between the HRFs and the LRFs with respect to 14 different characteristic, background, and satisfaction variables; and F-test analyses to determine if there were significant living environment differences between the HRFs and the LRFs on each of the subscales of the University Residence Environment Scale, Form R. The conclusions of the study were: 1. Neither high school nor college grades, SAT scores, nor individual characteristics, background, and satisfaction levels can be used to define differences in pledge retention between the HRF and LRF houses. 2. The differences within the living environment, and primarily the relationship dimension of that environment, of the two groups offer the best explanation of the pledge retention differences between the two groups. 3. The overt pressure exerted by the LRFs on their pledges to study and achieve academic success did not result in greater academic success than in HRFs, but did tend to limit the degree of social integration achieved by their pledges. 4. Social integration has a significant positive impact on pledge retention, while overt pressure toward academic integration has a probable negative impact on pledge retention. 5. Successful social integration, while having a positive impact on pledge retention, does not have a negative impact on academic performance. In fact, the impact on academic performance may be positive. 6. A crucial element in the Tinto (1987) model should be a relationship building block within the peer group interactin portion of the social system. Recommendations for further study were made. / Graduation date: 1991