The kidney is the second most frequent target of serious adverse effects of non-steroidal antiinflammatory drugs (NSAIDs). The renal side effects of NSAIDs related to inhibition of cyclooxygenase (COX) comprise reduction in renal blood flow (RBF) and glomerular filtration rate (GFR), sodium/water retention, water intoxication and hyperkalemia. The discovery of two COX-isoenzymes, a constitutive COX-1, serving homeostatic prostanoid synthesis, and an inducible COX-2, responsible for proinflammatory prosta noid production, led to the development of new NSAIDs: Preferential and specific COX-2 inhibitors, promising minimal NSAID-typical toxicity with equivalent efficacy. However, we learned that there is no clear distinction in "physiologic" constitutive COX-1 and "inflammatory" inducible COX-2. This is particular true for the kidney of humans and other mammalians, where COX-2 was found constitutively in meaningful amounts. Animal experiments and clinical trials with preferential and specific COX-2 inhibitors revealed that COX-2 is the critical enzyme for sodium excretion, renin release and likely antagonism of antidiuretic hormone. Additionally, a significant role of COX-2 for nephro genesis is suggested. For renal hemodynamics the given evidence point to COX-1 as the predominant enzyme, but further investigations are required. In summary, the gain of renal safety by use of preferential or specific COX-2 inhibitors is small or negligible with respect to sodium retention, hyperkalemia and probably water intoxication. These drugs may be advantageous regarding renal perfusion, but presently the same precautions as for conventional NSAIDs must be used.