A focus detection apparatus detects a first subject region (e.g., face region) from an image signal, detects a second subject region (e.g., human body region), in which the first subject region can be estimated, from the image signal, and performs focus detection based on an image signal corresponding to the set focus detection area. In the case where the primary subject corresponds to the first subject region detected, the focus detection area is set to a first region, and, in the case where the primary subject corresponds to the first subject region estimated based on the second subject region, the focus detection area is set to a second region that is larger than the first region.