The implication of credit rationing models states that the effect of monetary policy on output may be stronger when credit conditions are tight than when they are loose. Therefore, there may be a thresholde effect on the relation between real money supply and output. Existing empirical studies on testing threshold effects ignore the fact that the monetary policy and the credit conditions are endogenous, which are follow some optimal rules. Seeing that the past studies considering the endogenous monetary policy only cannot provide substantial evidence of the credit rationing theory, this article provides an extending test of threshold effects when monetary policy and the indicator of credit conditions are endogenous. Moreover, this study finds that the US aggregate data can still provide significant evidence of a threshold effect on the relation between money and output, comparing to the endogenous monetary policy partially considered.