Spectral imaging camouflage jamming tests were carried out with a kind of broad-band camouflage net and a hyperspectral imager. Based on the change of spectral characteristics of the spectral image data after being jammed, a spectral-distance criterion for evaluating jamming effect on spectral imaging was proposed by using spectral similarity measure. The applicability of the criterion was verified and analyzed with the data of the camouflage jamming tests. The results show that, under the test conditions, in the areas those are camouflaged effectively, the spectral distance between the spectral vectors of the image units before and after being camouflaged can reach 0.14~0.5 . The amplitude value varies with the change of the target characteristics before being camouflaged, camouflage net layers and the gain of the hyperspectral imager. On the other hand, the spectral distance of the image units remains below 0.05 in the non-camouflaged areas or the areas those are camouflaged ineffectively. Thus the spectral distance of the image units in camouflaged areas is usually larger than that of the image units in non-camouflaged areas remarkablely when the target is camouflaged effectively. The more remarkable the camouflage effect, the larger the spectral distance. Therefore the spectral distance between the spectral vectors of the image unit before and after being jammed can reflect the jamming effect on the spectral characteristics of the image unit quantitatively, and discriminate small differences in jamming effect on spectral imaging reliably. So the spectral distance could be used as an index for evaluating jamming effect on spectral imaging quantitatively and impersonally.