This paper analyzes the effects of antidumping and countervailing duty cases initiated from 1990 to 1997 that ended in withdrawn petitions without a suspension agreement or voluntary export restraint. Monthly import data are used to estimate the price and quantity effects of the withdrawn cases. The effects of the petition being withdrawn do not support the accepted wisdom that withdrawn petitions are a signal of collusion. However, a few of the cases show changes in price and quantities consistent with collusion. This is an important issue, since out-of-court settlements of unfair trade cases which restrict quantities or increase prices are not only welfare reducing but are also actionable under the antitrust laws; they are not exempt under the Noerr-Pennington doctrine.