The purpose of the present study was to determine the effect on littering behavior of (1) two ways of antilitter signs and (2) the condition of the environment (littered vs. unlittered). It was hypothesized that a sign with a threatening message (i.e. Littering is Unlawful and Subject to a $10 Fine.) would induce psychological reactance and therefore would be less effective than one emphasizing cooperation (i.e. Pitch In). It was also predicted that littering would occur more frequently in littered than in nonlittered areas. the study was conducted on six levels of a city parking garage, each floor of which was exposed to a randomly selected combination of sign and litter treatments. Relative to a no sign control condition, signs had the overall effect of reliably reducing the litter rate, but the "Pitch In" message was not found to be more effective than the Unlawful one. The littering rate was, as predicted, lowest in a clean environment. Finally, the impact of the signs was reliably influenced by the day of the observation. This finding was interpreted in terms of reactance theory and led to the conclusion that prior exposure is an important variable determining the effectivenss of signs.