This paper reports on an experiment designed to provide evidence of whether a change in the level of uncertainty about potential deductions influences the amounts that taxpayers take as deductions, and whether the effects of this change depend on the magnitude of the costs associated with defending the deductions, if challenged by the taxing authority. The results indicate that when the degree of uncertainty surrounding a deduction increases, taxpayers sometimes become more conservative in their choices, but at other times become more aggressive. In the early and middle rounds of the experiment, greater uncertainty resulted in lower deductions. However, in the late rounds this result reversed, with greater uncertainty leading to more aggressive reporting overall. In addition, this effect was much stronger when the cost of being audited was high. On one hand, these results are consistent with prior research in which experiments using larger penalties tend to provide stronger support for theoretical predictions. On the other hand, the aggressive reporting in response to increases in uncertainty documented in this experiment has not been demonstrated in other tax-reporting work, and likely suggests uncertainty does not have a clear, uniform effect on taxpayer behavior, but rather its effect depends on the specific nature of the decision-problem.