New tools are now available to provide a rigorous and systematic play-based exploration approach to the evaluation of unconventional resources. Coupled with petroleum system modeling, this methodology offers an efficient and effective approach to identify "sweet spots" early in the life of resource plays. Petroleum system modeling can be applied to predict the type and quantity of hydrocarbon in shale formations, as well as the proportion of adsorbed gas and geomechanical properties that are important for hydraulic fracture stimulation of shale reservoirs. Maps of these properties are then converted to chanceof-success maps for hydrocarbon generation, retention, and pore volume that can be integrated with nongeological factors, such as access and drilling depth required to reach target reservoirs. These play-based maps are expressed in probability units, so simple map multiplication provides a map of the play's overall chance of success, delineating the sweet spots. A similar methodology is applicable to evaluation of coalbed methane resources.In this paper, we illustrate this methodology using examples from shale oil and gas shale plays in North America. These include data-rich plays from the North Slope of Alaska and data-poor plays from the northeastern and southern regions of the United States, which are more representative of many Asia-Pacific basins. We show how predictions from petroleum system modeling based on sparse data provide a good match with results of subsequent development drilling and production.Petroleum system-based assessment of resources in place, combined with an assessment of overall play risk, enables companies to make decisions on acquisition of acreage early in the life of unconventional resource plays based on the probability of them containing economically viable resources.