The traditional argument that shorter product cycles favor trade secrets over patenting is based on the presumption that copying becomes less protable and trade secrets less likely to be discovered when a technology can be used for a shortened time period only. I review this argument. Under certain circumstances shorter product cycles can generate incentives to apply for more patents and the consequence may be the emergence of a patent thicket. In the model the underlying mechanism for the more intensive patenting activity is that one rm may start to raise its number of patent applications for exogenous reasons and the optimal reaction of other rms is then to match this behavior. Otherwise, the extensive patenting of one rm would drive the other rms out of the market or cut their prots signicantly. In order to prevent this, the other rms start ling patent applications on many ideas, which are not mature yet but may turn out successful eventually, instead of few fully developed technologies. If this situation occurs the firms may end up in a prisoners' dilemma where they would jointly prefer the situation where all firms innovate at moderate pace but each individual rm has an incentive to deviate to a short cycle with intensive patenting. Further, network effects may reinforce a firm's incentive to accelerate her R&D process and to induce other firms to adopt this strategy. Network eeffcts make it more attractive to gain market shares at an early stage because an advance compared to the competitors will be persistent. Therefore, the model predicts an intensive race by means of patent applications followed by a period of attenuated innovative activities. Similarly, blocking patents provide an advantage over competitors by limiting their access to technologies. But, in contrast to network effects, this is socially harmful because some R&D effort is wasted. Therefore, firms will patent less because their R&D tends to be less efficient. Licensing is an appropriate remedy to blocking patents and patent thickets and policy should facilitate licensing agreements in order to benefit from technologies that would otherwise lie idle.