Revision total knee arthroplasty (TKA) in the setting of bone deficiency requires varied levels of constraint to restore knee stability. However, the outcomes between different levels remain controversial. Clinical outcomes for 183 AORI Type I knees, 168 Type II knees and 124 Type III knees utilizing posterior stabilized (PS), unlinked constrained (UC) or hinged prostheses were evaluated with standardized clinical assessment tools and radiographic results over an average of 7.4 years. PS yielded superior knee scores in AORI Type I patients (P < 0.05), UC in Type II and III aseptic patients (P < 0.05), and a hinge was preferred in septic Type II or III knees (P < 0.05). Revision TKA conducted with increased constraint appears effective in the setting of increased bone deficiency.