Theories of optimal information sampling assert that the decision to gather information should take into account accrual costs: time, energy, and money. This project explored how effectively people traded off the monetary costs of gathering information against potential reward in a visuo-motor pointing task. We found that while subjects correctly adapt to changes in the cost structure of the task, almost all of them systematically sample more information than dictated by the optimal sampling rule (i.e., they over-sampled information relative to ideal).