AbstractPeak intensities of Fe Kα and scattered W Lα were repeatedly measured on 10‐μg iron‐impregnated and blank pieces of filter‐paper, repeatedly remounted or as mounted, at their 20 angles reset or as set, by the fixed‐time method at various times. Relative standard deviations (RSDs) of the Fe Kα intensity and Fe Kα/W Lα intensity ratio and the detection limit (DL) of iron were compared among these repeated measurements. With a constant total counting time, subdivision of the total time into a large number was preferable for the trace analysis of impregnated thin samples. Slight variations in the sample mounting had a large effect on the RSD and DL, whereas variation in the angle setting had a negligible effect. The use of the scattered W Lα intensity as an internal standard had no advantage in the trace analysis o