BACKGROUND There has been a general trend towards shortened length of post-kidney transplant hospitalization (LOH). The decision regarding patients's discharge from the hospital theoretically may be based on several factors, including, but not limited to, patient well being, insurance status, family situation and other, mostly socio-economic factors, as opposed to hard medical evidence. However, the appropriate LOH in kidney transplant recipients is not well studied regarding long-term outcomes. METHODS This study retrospectively analysed the association between LOH and graft and recipient survival based on United States Renal Data System dataset. In total, 100,762 patients who underwent transplant during 1995-2002 were included. Kaplan-Meier survival analysis and Cox models were applied to the whole patient cohort and on sub-groups stratified by the presence of delayed graft function, patient comorbidity index and donor type (deceased or living). RESULTS In recipient survival, both short (<4 d) and long (>5 d) LOH showed a significant adverse effect (p < 0.01) on survival times. In the analysis of graft survival, long LOH (>or=2 wk) also showed significant adverse effects (p < 0.001) on survival times. However, short LOH (<4 d) did not reach statistical significance, although it was still associated with adverse effects on graft survival. These observations were consistent across the whole patient cohort and sub-groups stratified by the presence of delayed graft function, patient comorbidity index and donor type. CONCLUSIONS Clinical considerations should be used to make the decision regarding appropriate time of post-kidney transplant recipient discharge. Based on this study, shorter than four d post-kidney transplant hospitalization may potentially be harmful to long-term graft and recipient survival.