This study was performed to clarify whether lifetime cadmium (Cd) intake (LCd) calculated from consumption of Cd-polluted rice and other foods, or drinking and/or cooking with Jinzu River water is associated with the occurrence of renal tubular dysfunction and to estimate the threshold level of LCd using a benchmark dose procedure. From participants in the 1967 and 1968 health examinations, the 2607 subjects (1430 men and 1177 women) who were aged 50 years or more, and who had either resided in the current hamlet since birth or moved there from a nonpolluted area and resided for a total of 30 years or longer were chosen as the target population. Based on a multiple logistic regression analysis performed for abnormal urinary findings (proteinuria, glucosuria, or proteinuria with glucosuria) using LCd as the explanatory variable with or without consideration of use of the river water for drinking and/or cooking, it was inferred that both factors of LCd and use of the river water influenced the occurrence of renal tubular dysfunction, with LCd appearing to exert a greater impact compared to drinking/cooking with the river water. Moreover, the threshold values of LCd for proteinuria with glucosuria were estimated to be 2.91 g in men and 1.79 g in women, without adjustment for use of the river water. After adjustment for its use, the values of proteinuria were calculated to be 0.93 g in women. When assuming user/nonuser status of the river water, the threshold level was calculated as 0.75/0.93 g in women. The difference of 0.18 g was interpreted as representing the influence of drinking/cooking with the river water. (C) 2008 Wiley Periodicals, Inc. Environ Toxicol 24: 421-428,2009.