Group A consisted of 35 literate young prisoners from Exeter's young prisoner wing. Group B consisted of 56 trainees from a young female's borstal. Group C was a consecutive sample of 45 adult male prisoners serving at least 18 months at Exeter prison. Group D consisted of 104 adult male prisoners serving at least 6 months at Exeter prison. One instrument had 10 items regarding perceptions of staff; the other had 20 items. Results for both scales confirmed the study hypothesis that some inmates experience staff as callous tyrants and other inmates see staff as treating inmates with fairness and respect. Two staff members rated the members of Group A. Prisoners who viewed staff as arbitrary tyrants were likely to behave in a resentful and difficult way. Equally, staff who view an inmate as being antiauthority may treat the inmate more harshly. Disciplinary reports for Group B showed correlations similar to those for the staff ratings. The differences in inmates' experiences of the same prison regime call into question the meaning that can be attached to average ratings. These ratings do not measure shared perceptions of the custodial environment, although an average can be used as one item in characterizing inmate experiences. However, other parameters, such as standard deviation, are as important. The effects of inmate personality on inmate perceptions should also be studied. Data tables, 11 references, and appendixes presenting study instruments are included.