OBJECTIVE:To assess whether a self-directed, computer-guided meditation training program is useful for stress reduction in hospital nurses. DESIGN: We prospectively evaluated participants before and after a month-long meditation program. The meditation program consisted of 15 computer sessions that used biofeedback to reinforce training. Participants were instructed to practice the intervention for 30 minutes per session, four times a week, for four weeks. Visual analogue scales were used to measure stress, anxiety, and quality of life (assessments were performed using Linear Analogue Self-Assessment [LASA], State Trait Anxiety Inventory [STAI], and Short-Form 36 [SF-36] questionnaires). Differences in scores from baseline to the study's end were compared using the paired t test. RESULTS: Eleven registered nurses not previously engaged in meditation were enrolled; eight completed the study. Intent-to-treat analysis showed significant improvement in stress management, as measured by SF-36 vitality subscale (P = .04), STAI (P = .03), LASA stress (P = .01), and LASA anxiety (P = .01). Nurses were highly satisfied with the meditation program, rating it 8.6 out of 10. CONCLUSIONS: The results of this pilot study suggest the feasibility and efficacy of a biofeedback-assisted, self-directed, meditation training program to help hospital nurses reduce their stress and anxiety. Optimal frequency of use of the program, as well as the duration of effects, should be addressed in future studies.