The tetragonal magnetostriction constant, (3/2)λ100, of Fe–Si single crystals was measured and was found to be structure dependent. Similar to that of Fe–Ge single crystals, (3/2)λ100 is positive in the single phase A2 regime, becomes negative in the single phase D03 regime, and changes from positive to negative between the two regimes. Short-range order in the A2 regime decreases the magnetostriction prior to the onset of long range order. In the single phase regions of both A2 and D03, thermal history does not show any obvious effect on the magnetostriction, contrary to that found for Fe–Ga alloys. However, in the regions of phase mixture involving A2, B2, and D03 phases, quenching pushes the change in magnetostriction from positive to negative to higher Si contents.