The choice between a set of alternatives often depends on how those alternatives are described, as well as their actual economic costs and benefits. We report results from an experiment designed to evaluate the impact of different descriptions of the after-tax wage on both (1) subjects’ willingness to perform a work task rather than an alternative leisure option, and (2) the amount of work performed by those subjects selecting the work task. Utilizing an experimental design that facilitates both within and between-subject comparisons, we find that that subjects’ willingness to work varies with the framing of the after-tax wage and that, in particular, subjects are much less willing to work when the returns to work are framed as a low wage plus a bonus than when the returns are described as a high wage minus a tax. Along the intensive margin we find suggestive evidence that subjects stop working just before their wage becomes subject to a significantly higher marginal tax rate, but we do not observe similar clustering when gross wages become subject to an equivalent wage decrease that is not described as a tax increase.