An ink jet recording apparatus and method for forming an image including a text region and a halftone region. The text region contains at least a part of a character and/or a thin line thinner than a predetermined thickness, and the halftone region contains a dot image on a recording medium. An image is formed by ejecting ink from recording heads in response to an input signal. The apparatus or method identifies the text region and the halftone region of an image on the basis of the input signal, expels black ink onto a recording area on the recording medium so as to form at least a part of the text region in a first scanning of the recording heads, expels a plurality of inks including the black ink onto the recording area so as to form at least a part of the text and the halftone regions in a second scanning, and repeating the first scanning and the second scanning on the next recording area after incrementing the recording medium.