Abstract: Neutron interactions, the neutron exchanges among two or more fissile systems, make the nuclear criticality safety problems very complicated in multiple units system. Unplanned neutron interaction between two fissionable bodies is involved in two criticality accidents. In recent years, the pressure of production capacity for nuclear fuels is growing rapidly with the growth of nuclear energy. A lot of multiple units systems are emerging in nuclear fuel manufacturing plant. The criticality safety of multiple unit systems is more and more important in nuclear fuel cycle. Towards the increasing specific applications of multiple units in fields of nuclear fuel cycle, a conservative, convenient and applicable assessment method was proposed to assess the nuclear criticality safety (reactivity changes) in multiple units system. The assessment method based on solid angle theory, the primary relations from distance factor to reactivity changes were established for double units with simple geometries, such as slabs, cylinders, and spheres. The reactivity change in one fissile unit from another one was treated as the neutron leakage change, which was quantitatively linked with the solid angle changes from other fissile units. The solid angle could be calculated by surface integral method. Then application scenarios for the assessment method were investigated and summarized to build up the numerical validation plans, such as nuclear fuel types, enrichments, container shapes, distance and shielding between containers. Double slabs, double cylinders and double spheres with 3 different ²³⁵U enrichments in uranium metal, uranium dioxide and uranyl nitrate systems were analyzed with Monte Carlo criticality calculation code. The contributions of neutron scattering between single units in slab system are stronger than other cases. This situation is also notable in shielding. Assessment method will not satisfy conservative condition. It's corrected by considering secondly scattering of leakage neutrons, through setting appropriate albedo conditions to adjust the solid angle values. The effects of materials and thickness of container wall were considered by container wall attenuation functions. Criticality experiments were performed to validate the assessment method on criticality assembly in China Institute of Atomic Energy. The criticality assembly is consisted of two parallel slabs with uranyl nitrate solutions with ²³⁵U enrichment 19.75%. The distance effect experiments include six different distances. The shielding effect experiments include three different types of shielding materials. The maximum bias between assessments and measurements is less than 0.12 Δk/k in distance effect experiments. The maximum bias between assessments and measurements is less than 0.035 Δk/k in shielding effect experiments. The feasibility and conservation of assessment method could be validated by the comparison results.