I will give a personal view of how modelling can best be used to assist public health policymakers in planning for and responding to emerging infectious disease threats. Staff at the MRC Centre at Imperial College have worked with policymakers around the world on a wide range of outbreaks, from Foot and Mouth Disease in UK cattle in the UK in 2001, to H1N1 'swine flu' in 2009. I will initially discuss how modelling has been used to assist in preparing for disease outbreaks, notably pandemic influenza, and the challenges of estimating the likely population impact of public health interventions (such as vaccines, antiviral treatment, school closure and other 'social distancing' measures) from limited data. Of particular note is the ability of modelling to give insight into the potential impact of combined - or layered - interventions of different types. Targeted layered interventions are now the mainstay of community mitigation planning for many developed countries, and modelling has therefore played an important role in defining pandemic plans. Giving examples from animal disease outbreaks, SARS in 2003 and pandemic influenza in 2009, I will then discuss the role of modelling in outbreak response - in giving real-time assessment of an emerging outbreak (notably assessing severity and speed of spread), generating projections of epidemic trajectory and informing decision-making on appropriate and effective control measures. I will discuss the difficulties faced in assessing severity and predicting the spread of the H1N1 pandemic last year and the wider challenges of real-time outbreak analysis, such as working with ever-changing and incomplete data, and needing to draw preliminary conclusions when underlying uncertainty is huge.