It was hypothesized that, with increasing length of police service, subjects would report fewer and less intense involvements in civilian community groups and fewer civilian friends, associates, and confidants. They would obtain higher alienation scores and would evidence less cognitive complexity in their perceptions of various community figures. Analysis of the data involved one-way analyses of variance with tests for trends and an examination of the important summary scores of four groups based on length of service. Pearson correlations were used to test the interaction hypotheses. Some of the findings reveal that (1) little differences appear in civilian group activity over time, but significant increases occur in the number and intensity of police group memberships; (2) a distinct tendency to replace civilians' personal relationships with police personal relationships emerges; (3) increased rather than decreased cognitive complexity occurs with increased length of service; and (4) subjects with the richest civilian group and informal relationships have lower alienation and cognitive complexity scores than those with more numerous attachments to police groups and associates. Findings indicate that the police officer does change over time. This change is most marked between police cadets and the least senior officers and may be attributed to the action of role-related factors which influence patterns of association and attitude. Seventeen data tables, survey instruments and 60 references are included.