To provide a speaker diaphragm which has high adhesive strength between the speaker diaphragm and a voice coil bobbin and has excellent reproduced sound quality less in peak and dip on a sound pressure frequency characteristic, and to provide a speaker.

The speaker diaphragm is molded by impregnating a thermosetting resin onto a base material. The speaker diaphragm is provided with an annular mounting part which is molded when the thermosetting resin is cured and provided projectingly toward a back of the speaker diaphragm with a cross-section shape changing on its circle. The annular mounting part respectively has a plurality of first mounting parts defined by a cross-sectional shape having an annular groove with which one edge of the voice coil bobbin is brought into contact and second mounting parts defined by a cross-sectional shape different from that of the first mounting parts, wherein the first mounting parts and the second mounting parts are arranged so as to alternately appear on one circle of the annular groove, the first mounting parts with different lengths L1 in an annular direction appear irregularly, and the second mounting parts with different lengths L2 in an annular direction appear irregularly.