Empirical work based on portfolio theories of the demand for money traditionally treat the quantity of money demanded as the outcome of a decision to allocate a fixed amount of wealth. By contrast, this paper argues that people commonly hold money at the same time that they borrow. Thus changes in gross wealth will cause changes in the quantity of money people wish to hold. This suggests that empirical work in the demand for broad money should pay attention to the cost of credit. We estimate the demand for broad money in the UK (1969Q1-1994Q4) using the recently developed time-series techniques of DOLS and FPLS. Our findings show that the cost of credit variables are important determinants of the demand for real M4 in the UK. In contrast to the recent empirical literature, our results provide strong evidence for the hypothesis that interest rates have a short and long term causal impact on real money demand in the UK.