OBJECTIVES: Demand for a primary TKA is increasing and projected to grow exponentially by 2030.  Comorbidities have been identified as a contributor to increased postoperative length of the stay (LOS) of the patient in the hospital, a major contributor to the cost of TKA. The purpose of the study is to assess the effect of comorbidities [diabetes, hypertension, obesity (bmi >30 kg/m2), chronic back pain(CBP), insomnia, cancer, osteoporosis], on patient LOS after elective TKA. METHODS: Patients ≥40 years old with a primary unilateral TKA, excluding those with a 2nd  TKA surgery within the year, were retrospectively extracted from a clinical orthopedic database between September 1, 2008 to November 30, 2010. LOS and comorbidities were obtained by chart abstraction from the electronic medical record. Unadjusted and adjusted odds ratios (OR) of having a LOS >3 days were calculated for each comorbidity.  Using logistic regression analysis, ORs were adjusted for age, gender, pre-operative physical/mental health and preoperative level of physical activity. RESULTS: A total of 174 patients with an average age of 63.8 (SD 10.2, 41-85) years were included in the study; 63% (110) were female and 56% (98) were obese.  75% of the patient population was discharged in ≤ 3 days.   Unadjusted analysis identified those with diabetes, hypertension, obesity, CBP, cancer and osteoporosis as being more likely to have a LOS >3 than those without these comorbidities. In adjusted ORs, the likelihood of having a LOS >3 days were 7.88 times (95% confidence interval[CI]=2.28-27.21) among patients with osteoporosis and 1.87 times  (95% CI =1.20-9.96) among patients with a history of cancer versus those without these comorbidities. CONCLUSIONS: Few studies have focused on specific comorbidities that influence LOS.   We found that preoperative diagnosis of osteoporosis or cancer increased the LOS following elective TKA. Previous determinants, age, gender, pre-operative level of physical activity, preoperative and physical health were not influential.