This paper explores the relationship between consumer confidence, stock prices and the business cycle in the United States using a Structural Vector Autoregression (SVAR). It finds three key results. First, the addition of confidence and stock price shocks to a small SVAR has important effects on the dynamic responses of the US economy. A confidence shock of four index points changes US GNP by 0.14% (noting that it is not uncommon for confidence shocks to total 20 points in a few consecutive quarters), while a 7% change in the S&P 500 leads to a 0.5% change in GNP. Second, the influence of these two shocks on the US business cycle in the second half of the twentieth century has been important at various times. Confidence shocks accounted for 19% of the total effect of structural shocks to GNP during the early 1990s recession, while stock prices contributed 20% of the effect of structural shocks to GNP in the 2001 recession. Finally, adding confidence and/or stock prices to the benchmark SVAR model leads to a small improvement in out-of-sample forecasting performance of GNP but this is not statistically significant. Nevertheless, confidence and stock prices do provide statistically significant incremental information during recessions.