The aim of this work was to study the influence of raw or retrograded enzyme resistant starch (enzyme resistant starch type 2 or 3) feeding to rats on the faecal and urinary excretion of N-nitrosoproline, which is a valuable index of endogenous nitrosation. Rats were either concurrently treated with 2,500 ppm dietary proline and 3,000 ppm nitrite in their drinking water or not treated. The diets contained 0% or 6% resistant starch type 2 or 3. No N-nitrosoproline was detected in the faeces and the urine of nontreated animals. Intake of raw and retrograded enzyme resistant starch by animals receiving nitrite and proline caused a decreased output of total N-nitrosoproline by 24% and 14%, respectively. On average, 69% of the N-nitrosoproline output was found in the urine and the remainder in the faeces. The observed reduction of the endogenous formation of N-nitrosoproline might be due to an increase of the viscosity of the contents of the stomach and the small intestine, thus lowering the formation and/or absorption of N-nitrosoproline. Whether the decrease of the caecal pH following enzyme resistant starch intake had any effect on the in vivo bacterial formation of N-nitrosoproline could not be demonstrated in this study. Further research is necessary in order to clarify this inhibitory effect of enzyme resistant starch feeding. (C) 1999 Elsevier Science Inc.