An image forming apparatus includes a pre-transfer conveying portion that conveys a recording medium, an image bearing member that moves while bearing a toner image, a conveying belt that attracts the recording medium conveyed from the pre-transfer conveying portion and conveys the recording medium at the same speed as the image bearing member, a detecting portion that detects the recording medium attracted to the conveying belt, and a control unit that controls the speed at which the pre-transfer conveying portion conveys the recording medium. On the basis of the detection of the recording medium, the control unit changes the speed of the pre-transfer conveying portion from a speed slower than the speed of the conveying belt to a speed equal to or faster than the speed of the conveying belt, and thereby causes the leading edges of the toner image and the recording medium to coincide.