Following an outbreak of highly pathogenic avian influenza virus (HPAIV) in a poultry house, control measures are put in place to prevent further spread. An essential part of the control measures based on the European Commission Avian Influenza Directive 2005/94/EC is the cleansing and disinfection (C&D) of infected premises. Cleansing and disinfection includes both preliminary and secondary C&D, and the dismantling of complex equipment during secondary C&D is also required, which is costly to the owner and also delays the secondary cleansing process, hence increasing the risk for onward spread. In this study, a quantitative risk assessment is presented to assess the risk of re-infection (recrudescence) occurring in an enriched colony-caged layer poultry house on restocking with chickens after different C&D scenarios. The risk is expressed as the number of restocked poultry houses expected before recrudescence occurs. Three C&D scenarios were considered, namely (i) preliminary C&D alone, (ii) preliminary C&D plus secondary C&D without dismantling and (iii) preliminary C&D plus secondary C&D with dismantling. The source-pathway-receptor framework was used to construct the model, and parameterisation was based on the three C&D scenarios. Two key operational variables in the model are (i) the time between depopulation of infected birds and restocking with new birds (TbDR) and (ii) the proportion of infected material that bypasses C&D, enabling virus to survive the process. Probability distributions were used to describe these two parameters for which there was recognised variability between premises in TbDR or uncertainty due to lack of information in the fraction of bypass. The risk assessment estimates that the median (95% credible intervals) number of repopulated poultry houses before recrudescence are 1.2 × 104 (50 to 2.8 × 106), 1.9 × 105 (780 to 5.7 × 107) and 1.1 × 106 (4.2 × 103 to 2.9 × 108) under C&D scenarios (i), (ii) and (iii), respectively. Thus for HPAIV in caged layers, undertaking secondary C&D without dismantling reduces the risk by 16-fold compared to preliminary C&D alone. Dismantling has an additional, although smaller, impact, reducing the risk by a further 6-fold and thus around 90-fold compared to preliminary C&D alone. On the basis of the 95% credible intervals, the model demonstrates the importance of secondary C&D (with or without dismantling) over preliminary C&D alone. However, the extra protection afforded by dismantling may not be cost beneficial in the context of reduced risk of onward spread.