致癌物質在高劑量告致癌是眾所周知的事實，惟並未有足夠的科學證據顯示暴露在極低劑量下會致癌，因此在高劑量所求得之劑量反應曲線模式要直接應用到低劑量的情況下是不太合理的，特別是在訂定環境中可容許之致癌物質標準時，因比如何訂定各種致癌物質之最低劑量標準使人們終其一生能免於其威脅乃一頗為棘手的問題，本研究旨在探討一合乎生理特性之劑量反應模式，此一模式考慮了(i)自然發生率(ii)生理代謝反應 (iii)疾病間競爭風險及(iv)致癌時間等特性與Weibull模式相結合。並對美國環保署訂定砷之標準所用之模式提出評論及推算環境中容許之飲用水質砷含量。

Carcinogenecity of a substance is frequently established at high doses. In the absence of hard medical evidence, some people estimate the risk of cancer at very low doses based on the “linear nonthreshold” theory which extrapolates the dose­response curve form high doses to low doses proportionally Many scientists question this theory when it is applied at levels of exposure so low that often neither the pollutant nor the health effect can be detected. The question arises as to the effects of the substance at the relatively low doses to which humans may be exposed in environment. Three classes of models (i.e. tolerance distribution models. mechanistic models, or time-to-tumor models) are usually employed as low-dose extraploation. A significant breakthrough in quantitative assesment was achieved by Gehring and Blau. (1977) who developed a simple yet pharmacokinetic model for carcinogenic process. To estimate safe dose, we attempt to pursue a plausible dose-response model which incorporates pharmacokinetic model, mechanistic model and time-to­tumor model together with considering competing risk factor. Also, the dose-response model adopted by the EPA of the United States to set a water quality standard for arsenic in America is examined.