The fundamental preservation problem in libraries is deciding what to preserve. A rather different approach to this decision making is presented. In class 1, for preservation, would go those items of economic value. Level 5 acquisitions (conspectus definition) also come into this coass not for their intrinsic value, but for the time and effort expended in collecting them. In class 2 would go items in high demand for curricula and research needs; replacement would be the technique for preservation adopted for high use materials. Class 3 material is ideal for co-operative preservation microfi.ming. Scholars in individual institutions it represents what consider of importance. No selection would be needed, but material in particular field not included would be at risk. Holding institutions would take the responsibility for preserving their collections according to a regional national survey of collections by subject. References and notes.