Nonlinear Residual Income Model Abstract Residual income has been proven to be a new approach of value relevance recently. The purpose of this study is to introduces residual income completely, and hopefully make some creativeness and contribution to residual income model. This paper is a both modeling and empirical study. In modeling, we have the following results¡G (1) Next period residual income is a nonlinear function of this period residual income, when we consider managers¡¦ real option. (2) This study introduces ¡§nonlinear residual income model¡¨ into Ohlson model, therefore firms¡¦ value is a nonlinear function of this period residual income. (3) This paper develops an option-based valuation model. According to this paper, equity value consists of the expected value from maintaining current operations, plus the value of the (put) option to discontinue operations at date t+1, and value of the (call) option to expand operations at date t+1. Empirical tests based on 27,536 firm-year observations from 1991-99 supports the above predictions of (1) and (2). In addition to the traditional OLS, this paper applies a new statistical approach--Sliced Inversed Regression (SIR). By SIR, we identify that our data has nonlinear components. This paper provides an alternative choice of valuation model and suggests that future research should approach the basic of value drivers.