A sheet processing apparatus which is capable of stabilizing a stop position of the sheet bundle, to thereby enable creation of a high finish quality product, and is compact in size and can be manufactured at low costs. Conveyance belts convey the sheet bundle to a flattening position. A pressing roller flattens the folded spine of the sheet bundle by pressing the folded spine. Clampers hold the sheet bundle such that the sheet bundle is not moved when flattening is performed. A CPU starts to stop driving of the conveyance belts when the sheet bundle is conveyed to a predetermined location upstream of the flattening position in the conveying direction of the sheet bundle, and stops the sheet bundle at the flattening position by holding the sheet bundle with the clampers before the conveyance of the sheet bundle is stopped.