Computation off-loading, i.e., remote execution, has been shown to be effective for extendingthe computational power and battery life of resource-restricteddevices, e.g., hand-held, wearable, and pervasive computers. Remote execution systemsmust predict the cost of executing both locally and remotely to determine when off-loading will be mostbeneficial. These costs however, are dependent upon the execution behavior of the task being considered and the highly-variable performance of the underlying resources, e.g.,CPU (local and remote), bandwidth, and network latency.As such, remote execution systems must employ sophisticated,prediction techniques that accurately guide computation off-loading. Moreover, these techniques must be efficient, i.e., theycannot consume significant resources, e.g., energy, execution time, etc.,since they are performed on the mobile device. In this paper, we present NWSLite, a computationally efficient,highly accurate prediction utility for mobile devices. NWSLite is an extension to the Network Weather Service (NWS),a dynamic forecasting toolkit for adaptive schedulingof high-performance Computational Grid applications. We significantlyscaled down the NWS to reduce its resource consumption yet stillachieve accuracy that exceeds that of extant remote executionprediction methods. We empirically analyze and compare both the prediction accuracy and the cost of NWSLite and a number of different forecasting methods from existing remoteexecution systems. We evaluate the efficacy of the differentmethods using a wide range of mobile applications and resources.