This paper estimates the effect of physical violence and property crimes on subjective well-being in Australia. Our methodology improves on previous contributions by (i) controlling for the endogeneity of victimisation and (ii) analysing the heterogeneous effect of victimisation along the whole distribution of well-being. Using fixed effects panel estimation, we find that both types of crimes reduce reported well-being to a large extent, with physical violence exerting a larger average effect than property crimes. Furthermore, using recently developed panel data quantile regression model with fixed effects, we show that the negative effects of both crimes are highly heterogeneous, with a monotonic decrease over the distribution of subjective well-being.