An in-vivo image capturing apparatus including a capsule casing, a gravity sensor, and an image selector. The capsule casing is insertable into a body of a subject and it includes an imaging optical system which can obtain images in the both end directions. The gravity sensor detects a gravity direction of the capsule casing within the body of the subject. The image selector selects one of the both end directions as an imaging direction of the imaging optical system, according to the detected result of the gravity direction by the gravity sensor.