Models and concepts of sandstone diagenesis developed over the past two decades are currently employed with variable success to predict reservoir quality in hydrocarbon exploration. Not all of these are equally supported by quantitative data, observations, and rigorous hypothesis testing. Simple plots of sandstone porosity versus extrinsic parameters such as current subsurface depth or temperature are commonly extrapolated but rarely yield accurate predictions for lithified sandstones. Calibrated numerical models that simulate compaction and quartz cementation, when linked to basin models, have proven successful in predicting sandstone porosity and permeability where sufficient analog information regarding sandstone texture, composition, and quartz surface area is available. Analysis of global, regional, and local data sets indicates the following regarding contemporary diagenetic models used to predict reservoir quality. (1) The effectiveness of grain coatings on quartz grains (e.g., chlorite, microquartz) as an inhibitor of quartz cementation is supported by abundant empirical data and recent experimental results. (2) Vertical effective stress, although a fundamental factor in compaction, cannot be used alone as an accurate predictor of porosity for lithified sandstones. (3) Secondary porosity related to dissolution of framework grains and/or cements is most commonly volumetrically minor (<2%). Exceptions are rare and not easily predicted with current models. (4) The hypothesis and widely held belief that hydrocarbon pore fluids suppress porosity loss due to quartz cementation is not supported by detailed data and does not represent a viable predictive model. (5) Heat-flow perturbations associated with allochthonous salt bodies can result in suppressed thermal exposure, thereby slowing the rate of quartz cementation in some subsalt sands.