Staphylococcus aureus (SA) colonises and can damage skin in atopic dermatitis (AD) lesions, despite being commonly found with Staphylococcus epidermidis (SE), a commensal that can inhibit SA’s virulence and kill SA. Here, we developed an in silico model, termed a “virtual skin site”, describing the dynamic interplay between SA, SE, and the skin barrier in AD lesions to investigate the mechanisms driving skin damage by SA and SE. We generated 106 virtual skin sites by varying model parameters to represent different skin physiologies and bacterial properties. In silico analysis revealed that virtual skin sites with no skin damage in the model were characterised by parameters representing stronger SA and SE growth attenuation compared to those with skin damage. This inspired a treatment strategy combining SA-killing with an enhanced SA-SE growth attenuation, which in silico simulations found recovers many more damaged virtual skin sites to a non-damaged state, compared to SA-killing alone. This study demonstrates that in silico modelling can help elucidate key factors driving skin damage caused by SA-SE colonisation in AD lesions and help propose strategies to control it, which we envision will contribute to the design of promising treatments for clinical studies.