The paper is dealing with the controversial question of the potential impact of stock market fluctuations on consumer confidence. In the last few years, this confidence index has gained importance in business cycle analysis and empirical evidence has shown its explanatory power in forecasting consumption along with standard macroeconomic variables. Meanwhile, numerous interpretations of its fluctuations arose, and few were based on a solid argumentation. Therefore, we propose in this paper to determine which elements are actually driving the confidence index. Using the standard error-correction mechanism model and non-linear methods, we analyze the relationship between the confidence index and several economic variables, over the period ranging from January 1983 to December 2001. As a growing number of economic observers claim the stock market fluctuations have a strong impact on consumer confidence, we especially focus on this potential impact. The models are estimated for the United States and for Belgium for which the importance of equities in the households net wealth is quite different. We find in particular that stock market fluctuations have explanatory power in the evolution of consumer confidence in the United States, especially since the beginning of the nineties.