This paper investigates the effect of correlation and volatilities of firm and project cash flows on the choice of project finance. I use a pure-play approach to measure unobservable project cash flows for a sample of 440 US and non-US firms that invested in 577 projects from 1990 to 2008 and find evidence that the probability of project finance is increasing in cash flow volatility difference between firm and project cash flows. The likelihood of the project finance is greater when volatilities are different and the correlation between firm and project cash flows is high. I also find that firms are likely to choose corporate finance for low correlation and low and similar volatilities between firm and project cash flows. This empirical work is consistent with the theoretical predictions in Leland (2007) that provides a potential explanation for the existence of project finance based on financial synergies.