Background: Atherosclerotic plaque progression is a complex process to which several physiological pathways contribute. Smoking has been shown to accelerate plaque growth by increasing inflammation, thrombosis, and endothelial cell (EC) dysfunction, but the relative contribution of each pathway is not well understood. Cellular adhesion molecules (CAMs) are preferentially produced by activated ECs and have therefore been proposed as a means to quantify EC dysfunction. Measurements of CAMs in smoke-exposed ApoE-/- mice, however, have produced ambiguous results, with both positive and negative changes relative to control. Objective: modeling the dynamics of EC membrane-bound inter-cellular adhesion molecule 1 (mICAM-1) will advance our understanding of the impact of cigarette smoke exposure on EC function in the ApoE-/- mouse and allow improved design of future non-clinical experiments. Methods: we licensed the Entelos ApoE-/- mouse cardiovascular Physiolab® platform, a comprehensive in silico modeling technology enabling dynamic simulation of the biological pathways driving plaque progression. Biological effects of smoking were incorporated into the model and calibrated to histological data on total plaque area and plaque percent macrophage area. The concentration of mICAM-1 was simulated with and without smoking effects to predict the time-dependent ratio of mICAM-1 in smoke-exposed mice and controls. An uncertainty analysis was conducted to generate alternative falsifiable predictions that could be tested with in vivo experiments. Results:The simulation results showed that, due to competing biological effects, mICAM-1 follows a bell-shaped curve as a function of time. Under simulated smoking conditions, the curve maximum occurs sooner than with non-smoking; consequently, the log-ratio of mICAM-1 in smoke-exposed and control mice may be positive or negative, depending on the time of assessment. The modeling results are supported by in vivo measurements of mICAM-1 in ApoE-/- mice collected using an ultrasound and micro-bubble technique. By leveraging the simulations, we were then able to propose a new experimental protocol that will improve our quantification of the impact of smoking on EC function. Conclusion: We have used an in silico model to mechanistically explain contradictory experimental data. Furthermore, such model predictions can be used to optimize protocol design for future in vivo studies.