The research rested on the premise that each individual responds differently to stressors, and each has personal characteristics that either reduce or reinforce stress. The study examined both personal characteristics and work-related stressors and measured stress by means of two measures: physiological symptoms and indicators of physical, emotional, or mental exhaustion. The subjects were randomly chosen from the rank of sergeant and below. Fifty-seven percent returned usable instruments. Officers did not describe themselves as highly stressed, perhaps partly because of the lack of awareness of job-related stressors. As stress scores increased, ratings of job satisfaction decreased. In addition, 26 percent reported high blood pressure; 19 percent, severe migraine headaches; 11 percent, ulcers; and 1 percent, cancer. Irritability in the last 6 months was reported by 45 percent, frequent colds by 37 percent, excessive eating or appetite loss by 36 percent, and sleep problems by 36 percent. Thus, sizable percentages were experiencing stress. A model based on physiological indicators was a better predictor of stress than was one based on physical, emotional, and mental problems. Agencies should focus on prevention of stress rather than only on treatment. Fair workload assignments, stricter selection and promotional criteria, and integration of employees into the agency's problemsolving and decisionmaking networks are recommended. 13 references.