Background Assessing severity and spread of a novel influenza strain at the start of a pandemic is critical for informing a targeted and proportional response. It requires community-level studies to estimate the burden of infection and disease. Rapidly initiating such studies in a pandemic is difficult. The study aims to establish an efficient system allowing real-time assessment of population susceptibility, spread of infection and clinical attack rates in the event of a pandemic. Methods We developed and appended additional survey questions and specimen collection to the Health Survey for England (HSE) – a large, annual, rolling nationally representative general population survey recruiting throughout the year – to enable rapid population-based surveys of influenza infection and disease during a pandemic. Using these surveys we can assess the spread of the virus geographically, by age and through time. The data generated can also provide denominators for national estimates of case fatality and hospitalisation rates.Phase 1: we compared retrospectively collected HSE illness rates during the first two infection waves of the 2009 pandemic with the Flu Watch study (a prospective community cohort). Monthly and seasonal age-specific rates of illness and proportion vaccinated were compared.Phase 2: we piloted blood specimen and data collection alongside the 2012–13 HSE. We are developing laboratory methods and protocols for real-time serological assays of a novel pandemic influenza virus using these specimens, and automated programmes for analysing and reporting illness and infection rates.Phase 3: during inter-pandemic years, the study enters a holding phase, where it is included in the yearly HSE ethics application and planning procedures, allowing rapid triggering in a pandemic.Phase 4: once retriggered, the study will utilise the methods developed in phase 2 to monitor the severity and spread of the pandemic in real time. Results Phase 1: the rates of reported illness during the first two waves in the HSE underestimated the community burden as measured by Flu Watch, but the patterns of illness by age and time were broadly comparable. The extent of underestimation was greatest for HSE participants interviewed later in the year compared with those interviewed closer to the pandemic. Vaccine uptake in the HSE study was comparable to independent national estimates and the Flu Watch study.Phases 2 and 3: illness data and serological samples from 2018 participants were collected in the 2012–13 HSE and transferred to the University College London Hospital. In the 2013 HSE and onwards, this project was included in the annual HSE ethics and planning rounds. Conclusions The HSE’s underestimation of illness rates during the first two waves of the pandemic is probably due to recall bias and the limitation of being able to report only one illness when multiple illnesses per season can occur. Changes to the illness questions (reporting only recent illnesses) should help minimise these issues. Additional prospective follow-up could improve measurement of disease incidence. The representative nature of the HSE allows accurate measurements of vaccine uptake. Study registration This study is registered as ISRCTN80214280. Funding This project was funded by the NIHR Public Health Research programme and will be published in full inPublic Health Research; Vol. 3, No. 6. See the NIHR Journals Library website for further project information.