Renal disease has a high incidence in cats, and some evidence implicates dietary P as well. To investigate this further, two studies in healthy adult cats were conducted. Study 1 (36 weeks) included forty-eight cats, stratified to control or test diets providing 1·2 or 4·8 g/1000 kcal (4184 kJ) P (0 or approximately 3·6 g/1000 kcal (4184 kJ) inorganic P, Ca:P 1·2, 0·6). Study 2 (29 weeks) included fifty cats, stratified to control or test diets, providing 1·3 or 3·6 g/1000 kcal (4184 kJ) P (0 or approximately 1·5 g/1000 kcal (4184 kJ) inorganic P, Ca:P 1·2, 0·9). Health markers, glomerular filtration rate (GFR) and mineral balance were measured regularly, with abdominal ultrasound. Study 1 was halted after 4 weeks as the test group GFR reduced by 0·4 (95 % CI 0·3, 0·5) ml/min per kg, and ultrasound revealed changes in renal echogenicity. In study 2, at week 28, no change in mean GFR was observed (P >0·05); however, altered renal echogenicity was detected in 36 % of test cats. In agreement with previous studies, feeding a diet with Ca:P <1·0, a high total and inorganic P inclusion resulted in loss of renal function and changes in echogenicity suggestive of renal pathology. Feeding a diet containing lower total and inorganic P with Ca:P close to 1·0 led to more subtle structural changes in a third of test cats; however, nephrolithiasis occurred in both diet groups, complicating data interpretation. We conclude that the no observed adverse effects level for total dietary P in adult cats is lower than 3·6 g/1000 kcal (4184 kJ), however the effect of inorganic P sources and Ca:P require further investigation.