Individual preferences for 'ageing in place' (AIP) in old age are not well understood. One way to test the strength of AIP preference is to investigate the effect of health shocks on residential mobility to smaller size or value dwellings, which we refer to as 'housing downsizing'. This paper exploits more than a decade worth of longitudinal data to study older people's housing decisions across a wide range of European countries. We estimate the effect of health shocks on the probability of different proxies for housing downsizing (residential mobility, differences in home value, home value to wealth ratio), considering the potential endogeneity of the health shock to examine the persistence of AIP preferences. Our findings suggest that consistently with the AIP hypothesis, every decade of life, the likelihood of downsizing decreases by two percentage points (pp). However, the experience of a health shock partially reverts such culturally embedded preference for AIP by a non- negligible magnitude on residential mobility (9pp increase after the onset of a degenerative illness, 9.3pp for other mental disorders and 6.5pp for ADL), home value to wealth ratio and the new dwelling's size (0.6 and 1.2 fewer rooms after the onset of a degenerative illness or a mental disorder). Such estimates are larger in northern and central European countries.