Statistical fluctuations, or noise, in reconstructed tomograms are analyzed as a function of x‐ray energy. The result is essentially constant for energies above 50 keV, assuming constant dose and spatial resolution, and is in close agreement with the actual performance of the EMI‐Scanner (water‐bag model). The noise is inversely proportional to the square of cell dimensions. For scanners without a water bag, the theoretical noise is approximately 30% less.