The irrigation efficiency paradox says that raising the efficiency of irrigation systems, thereby reducing return flows, either gives no change in water depletion or it raises depletion via increased evapotranspiration and irrigated area. While this paradox can occur, there are problems associated with it. It eludes precise explanation and characterisation; it can be confused with other irrigation hydrology paradoxes; it is one of several ways irrigated areas increase; it over-emphasises the role of return flows; it relies on other irrigation variables (usually unstated) being uncontrolled; it can be inverted to reduce depletion; and it may mistakenly guide the conservation of water in irrigated systems. Addressing these concerns, a comprehensive predictive model called Irrigated Systems Accounting (ISA) analyses irrigation undergoing water conservation based on accounts for soil-crop evapotranspiration, irrigation efficiency (IE), irrigation practices and infrastructure, withdrawals, depletion, crop production and water reallocation. By using more calculi than current water accounting, ISA; resolves irrigation efficiency paradoxes; predicts how an irrigated system changes its aggregate area and depletion via primary, expansion and reuse zones; and reveals how other non-IE factors drive up area but not necessarily depletion. Compiling all zonal changes reveals how reductions in aggregate depletion can be derived and reallocated to other users without cutting crop production. The paper concludes there are hazards for water policy if irrigation efficiency and depletion are exclusively tied together via imprecise characterisations that draw on water accounting models containing few terms and relationships.