The work in this thesis examines the effect of mean stress on the fatigue behaviour of very hard (Rockwell C 60) steels (AISI 8822, 8620, 9310, and cold-worked pre-stressing wire).In the mean stress tests, the minimum stress in the fatigue cycle was varied from test to test over a range from -1200 MPa to a value approaching the true fracture stress of each material.The results are not adequately explained by current theories for the effect of mean stress on fatigue behaviour in the region of compressive mean stresses. All current theories suggest that the maximum stress at the fatigue limit decreases with decreasing minimum stress. The results of this study shows that instead of continuing to decrease with decreasing minimum stress the maximum stress at the fatigue limit remains constant indicating an insensitivity to the minimum stress in the fatigue cycle for minimum stresses below the value in a fully reversed fatigue test. The theory proposed by the author corrects this error by maintaining the maximum stress at the fatigue limit constant with decreasing minimum stress in the region of negative mean stresses. The results are of interest to designers of components in which high negative residual stresses are introduced into materials hardened by, for example, carburizing, nitriding, or induction hardening to improve the fatigue strength of components. The present work allows considerably higher design stresses for operating stresses in the negative mean stress region than previous theories permit. v ACKNOWLEDGEMENTSFirst and foremost, I offer my most sincere and deepest gratitude to my MASc supervisor, Professor Emeritus Dr. Timothy Topper. Throughout my stay at the University of Waterloo, he guided me and taught me everything, from how to run fatigue tests in the laboratory, through helping me understand Metal Fatigue concepts, to how to adapt to the cold Winters in Canada. He provided encouragement, sound advice, great mentoring, and great company.With his endless patience and energy, he was always accessible and willing to help. One simply could not wish for a better or friendlier supervisor. I would also like to thank my co-supervisor, Dr. Grzegorz Glinka, who has showed his support in a number of ways and without whom I wouldn't be here. I had the honour to attend his class and he sets an example of a world-class researcher for his passion on knowledge.