There is provided an image capturing apparatus. An image capturing unit acquires a signal in a first acquisition mode or a second acquisition mode for each pixel of an image sensor. In the first acquisition mode, an image signal obtained by adding focus detection signals of a plurality of photoelectric conversion units of the pixel is acquired. In the second acquisition mode, the focus detection signals are acquired in addition to the image signal. The image capturing alternately captures a recording image and a focus detection image having a smaller number of pixels than the recording image, applies all pixels to the first acquisition mode when capturing the recording image, and applies at least a part of the pixels to the second acquisition mode when capturing the focus detection image.