This paper examines the reasons underlying the wide variations in abundance and geometries of fluid-filled spaces in limestone and dolostone petroleum reservoirs. The main factors commonly identified as controlling this include (1) depositional texture; (2) eogenetic processes (a wide spectrum of early changes not limited to mechanical compaction, dissolution of less stable components, neomorphism of micrite, and various types of cementation); (3) anhydrite cementation; (4) dolomitization; (5) overpressure; (6) oil emplacement; (7) content of clay and stylolites; (8) burial dissolution; (9) karsting; (10) fracturing; and (11) sequence stratigraphy. Factors 1–4, 6, 7, and 11 are illustrated by reviewing and comparing published results from studies of six carbonate platforms from the author’s personal experience. This new synthesis can provide a useful basis for developing realistic expectations for undrilled prospects and for explaining the wide diversity in observed carbonate porosity variations. Nevertheless, no new universal tools for porosity prediction are proposed, as each carbonate reservoir tends to be unique in the way that the above processes complexly interact to either amplify or diminish their influences on porosity evolution. Understanding will always ultimately rely upon reference to carefully observed empirical case studies.