N oise data are presented for various electron multiplication methods used in imaging devices. It is shown that the noise of these devices can always be represented in the same manner and that it can be described by a noise deterioration factor r, which is the ratio of the output noise over the multiplied input noise. In the transmission secondary electron multiplier, r decreases with increasing primary energy and with increasing collector voltage in such a manner that I’ is a universal function of the current multiplication factor s. Part of this may be due to electron losses in the aluminum backing, but the secondary emission multiplication also contributes. In the image intensifier multiplier, r decreases with increasing primary energy and shows comparable behavior to the first multiplication method. The gain here is frequency dependent because of the response of the phosphor. In the channel multiplier, the de gain Is/ /p decreases with increasing I‘ll’ so that the device is nonlinear. One must then distinguish between “apparent” noise deterioration factor r’ and the true deterioration factor r. It is found that F1 -cr, explaining the remarkably low value of r`(1.4) found in these devices.