The structure of an untwinned aenigmatite crystal from Vesterøya, Vestfold district, Norway, having a composition (Na3.73Ca0.27)∑=4.00 (Fe2+8.55Ti2.10Mg0.46Fe3+0.40Mn0.40Ca0.06Zn0.01Zr0.01)∑=11.99 (Si11.10Al0.64Fe3+0.26)∑=12.00 O40 was refined in space group P1̅ to give R1 = 0.0277 for 5515 unique reflections with Fo > 4σ(Fo) and 0.0324 for all 6145 unique data; GooF 1.124. Cell parameters are a = 10.415(1), b = 10.840(1), c = 8.931(1) Å, α = 105.107(4), β = 96.610(5), γ = 125.398(4)° and V = 746.8(1) Å3. Mössbauer spectroscopy of a bulk sample gave Fe3+/Fetot = 0.12(1) whatever was the initial fitting model, and IVFe3+/Fetot = 0.04(2), which is consistent with IVFe3+/Fetot = 0.018(1) obtained from occupancy of T3, the only T site to have Fe. Ti is fully ordered at M7, the site with the most shared edges, whereas VIFe3+ is mostly ordered at the smallest M sites (M1 and M2), consistent with what is reported in related minerals of the sapphirine group. Ca is ordered at M8, one of the two 8-coordinated sites. Bond lengths and bond-valence sums suggest that T2 and T4 are occupied by Si only, and that T1 and T3 have the highest proportion of Al. Ordering of Al3+ and Fe3+ at T3 could be related to Ca being ordered at M8, because M8 is linked with T3 through four oxygen anions, more than M8 with any other T site. Thus, one of the four substitutions proposed for aenigmatite, Ca + IVAl ⇔ Na + IVSi, could be operating at a local scale. Since (IVAl + IVFe3+) exceeds Ca, there must be other substitutions for incorporating Al and Fe3+. Two of the four proposed substitutions, VIFe2+ + IVSi ⇔ VIFe3+ + IVAl and VIFe2+ + IVSi ⇔ VIFe3+ + IVFe3+, are consistent with the compositional and Mössbauer data, but the Ti content of 2 per 28 cations rules out incorporation of VIFe3+ in a wilkinsonite component, Na4Fe2+8Fe3+4Si12O40.