Natural killer (NK) cells are a key part of the body’s innate immune system, killing cells that have been damaged or stressed by infection and controlling the spread of disease while a slower, more powerful adaptive immune response can be prepared. They accomplish this by the controlled release of potent mediators of cell death like granzyme B, as well as the pore-forming perforin molecules that allow them to pass through target cell membranes. NK cells’ ability to destroy other host cells has made them a focus of much research into fighting cancer and microbial infections; however, this same potential for harm also necessitates a delicate system of regulation that must be interacted with carefully to minimize collateral damage to healthy host tissue. This study seeks to understand the movement of the granzyme B and perforin released from an NK cell using COMSOL Multiphysics 5.5, and ultimately to assess and quantify nearby cellular death. Granzyme B and perforin movement, transformation, and accumulation were modeled by mass transfer physics that accounted for reactions, diffusion, and partitioning across cell membranes. We assumed a cubical computational domain containing spherical subdomains: an NK cell, a target cell, and bystander cells. Some key model parameters are the amount of granzyme B and perforin released, their diffusion rates in different domains, their degradation rate constants, and the minimum lethal quantity of granzyme B. Damage to the cell was defined as reaching a threshold of granzyme B accumulation concentration inducing apoptosis, or cell death. The current literature suggests that extracellular concentrations in the picomolar range are sufficient to cause apoptosis [1], from which an estimate of lethal intracellular concentration was derived. We used two different models, one in which the NK cell attaches itself to a damaged cell and releases toxins asymmetrically, and another in which the NK cell is free-floating, and releases toxin uniformly across its surface. For each model, we randomized the cell locations and averaged the effects found in each scenario. This random distribution reflects the unordered arrangement of cells in vivo and provided a more refined perspective into bystander cell death. Modulation of NK cell activation can help us understand diffusion from a non-convergent release and its feasibility in clearing pathological cells while minimizing damage to healthy cells. Using the non-converged model, we found the concentration of toxin throughout the system at large and calculated cell death. We found that, across all five configurations of cells, less than 5% of the bystander cells were killed by the NK cell and the target cell was always killed. We thus concluded that the NK cell mechanism, even for non-converged release, works as intended and kills its target with little collateral damage. We were also able to discover important mechanisms in this system with relevance to genetic engineering of NK cells, perforin, and granzyme B.We found that the molecule count of granzyme B and perforin was highly effective in changing the concentration of internalized granzyme B, especially in comparision to changing the diffusivity through the extracellular fluid. A higher granzyme B diffusivity resulted in reduced granzyme internalization, which allows us to conclude that the two species reaching the cells at the same time is more important for their function than for one species to move to the cells very quickly.