The effects of the Transcendental Meditation program in a maximum security prison were studied via cross-validation design. Each of two experiments involved the pre- and postmeasurement of a treatment and a control group with a 14-week treatment interval. All subjects were administered the Eysenck Personality Inventory, the State-Trait Anxiety Inventory, the Buss-Durkee Hostility Inventory, and a survey of sleep patterns and smoking habits. Parallel significant differences between the Transcendental Meditation and control groups across all the inventories were found via MANCOVAs, indicating reduction in anxiety, neuroticism, hostility, and insomnia as a function of the treatment.