BACKGROUND: Avian influenza (AI) is a public health challenge because of ongoing spread and pandemic potential. Non-pharmaceutical measures are important to prevent the spread of AI and to contain a pandemic. The effectiveness of such measures is largely dependent on the behaviour of the population. Risk perception is a central element in changing behaviour. This study aimed to investigate perceived vulnerability, severity and precautionary behaviour related to AI in the Netherlands during seven consecutive surveys in 2006 - 2007 as well as possible trends in risk perception and self-reported precautionary behaviours. METHODS: Seven web-based surveys were conducted including 3,840 respondents over a one-year period. Time trends were analyzed with linear regression analyses. Multivariate analysis was used to study determinants of precautionary behaviour. RESULTS: While infection with AI was considered a very severe health problem with mean score of 4.57 (scale 1 - 5); perceived vulnerability was much lower, with a mean score of 1.69. While perceived severity remained high, perceived vulnerability decreased slightly during a one-year period covering part of 2006 and 2007. Almost half of the respondents (46%) reported taking one or more preventive measures, with 36% reporting to have stayed away from (wild) birds or poultry. In multivariate logistic regression analysis the following factors were significantly associated with taking preventive measures: time of the survey, higher age, lower level of education, non-Dutch ethnicity, vaccinated against influenza, higher perceived severity, higher perceived vulnerability, higher self efficacy, lower level of knowledge, more information about AI, and thinking more about AI. Self efficacy was a stronger predictor of precautionary behaviour for those who never or seldom think about AI (OR 2.3, 95% CI 1.9 - 2.7), compared to those who think about AI more often (OR 1.5, 95% CI 1.2 - 1.9). CONCLUSIONS: The fact that perceived severity of AI appears to be high and remains so over time offers a good point of departure for more specific risk communications to promote precautionary actions. Such communications should aim at improving knowledge about the disease and preventive actions, and focus on perceived personal vulnerability and self efficacy in taking preventive measures.