Noise of automotive alternators can be classified into mechanical noise, aerodynamic noise and electro-magnetic noise. which is the same as for electric motors. Previous studies show that the elect ro-magnetic noise takes a maw peak at the rotating frequency multiplied by the number of stator slots. It has not been proved clearly so far, however, that the major peak is wholely due to the stator slots. On the contrary it is well known that noise of motors. which has a mechanism similar to the alternator except that the number of stator slots in automotive alternators is in geneinteger multiple of that of rotor segments, is closely related to the number of rotor slots. Therefore, the statement that only the stator slots is the source of the major peak in the noise spectrum of alternators is suspicious although not easy, to show theoretically, that the statement is incorrect. In this paper. effects of the stator slots on the noise in an automotive alternator are experimentally investigated by intentionally modifying the number of stator slots in such a way that the number of the states is not an integer multiples of the rotor slots. It is shown that both the stator slots are not so much influential as the rotor slots and claimed that the major peak in the noise spectrum of conventional alternators is due to superposition of a component caused by the stator and a higher harmonic component caused by the rotor