A piezoelectric crystal is a unit that changes its frequency in parallel with a change in its mass. This characteristic is exploited in designing flow cell-based immunosensors for detecting the concentration of antibodies in liquid samples. In the present study, computational fluid dynamic techniques are used to optimize the antigen-antibody binding process on an electrode surface placed on the base of a conical flow cell. The geometry optimization of the flow cell was determined to minimize the test time. This time is needed for the electrode to be saturated by the antibody, a process that requires the maximization of the adsorption rate and be accomplished by increasing the shear rate in the vicinity of the electrode. To validate the numerical model and to determine its parameters, experiments were carried out using an identical flow cell. In the experiments, the system did not reach saturation within an acceptable time frame, therefore, the model parameters were determined based on the unsaturated state. The experimental results confirmed the applicability of numerical simulations in predicting the effect of changing the inlet section area of the flow cells, proving the computational model to be very valuable in designing immunosensors based on flow cells.