Urban landscapes are important factors that affect housing prices, and significant differences between landscape preferences of various homebuyers may be observed because of the different reasons for purchasing a house (consumption or investment). However, the hedonic price model widely applied in most existing studies only captures the average effects of landscapes as a whole sample, and may ignore the heterogeneity of landscape preferences. To fill this gap, this study constructed hedonic price models and quantile regression models with the housing data in Guangzhou, China from 2013 to 2016 and analyzed the landscape preferences of buyers with different purchase motivations. Empirical results showed that the landscape preferences of buyers were different in housing submarkets. The implicit value of landscapes was greater in consumption demand than in investment demand, whereas investment buyers were more vulnerable to the disamenity effect of unattractive landscapes. In addition, the quantile effect of landscapes was identified, in which the buyers of highpriced housing will pay more for high-quality landscapes. This study revealed the diversified housing demands and landscape preferences of homebuyers, which is important for urban planning and project development.