BACKGROUND Aseptic implant loosening is one of the most common complications leading to revision surgery in total elbow arthroplasty. Different humeral stem lengths are available with varying designs. In general, the decision of which stem length to use depends on the surgical diagnosis or simply the surgeon preference. Often, the longer stem is used for post-traumatic or revision cases while for rheumatoid patients the shorter stem is preferred. There are no data in the literature to favor one humeral stem size over the other according to the diagnosis. METHODS We analyzed the total elbow joint database of the Coonrad-Morrey design at our institution for aseptic loosening leading to revision and compared the revision rate and the survival of the 4- and 6-inch humeral stems. RESULTS Overall, revision for aseptic humeral loosening is infrequent and occurred in only 16 of 711 total elbow arthroplasties during a mean follow-up of 88 months. There was no significant difference in the revision rate between the 2 stem lengths (1.9% for the 4-inch stems and 2.6% for the 6-inch stem). CONCLUSION Revision rate was correlated to the surgical diagnosis and was significantly higher for post-traumatic patients than for rheumatoid patients (5.1% vs 0.66%, P < .001). Of interest, and possibly not surprising, the mean time to revision was shorter for the 4-inch stems than it was for the 6-inch stems (37 vs 95 months, P = .034).