An image forming apparatus includes a first detector that detects an operation from which input of a start instruction for forming an image is predicted, a second detector that detects a human body, and a controller that controls starting a preparation operation even without input of the start instruction in a case where the operation is detected. The controller controls ending the preparation operation where the start instruction is not input even after a first predetermined time has passed since the start of the preparation operation and the second detector does not detect the human body, and controls continuing the preparation operation in a case where the start instruction is not input even after the first predetermined time has passed since the start of the preparation operation and the second detector detects the human body.