Metal complexes (M = Li+, Na+, K+, Rb+, Cs+, Ca2+, Sr2+, Ba2+, Ag+) of the two closely related pairs of macrocycles FN2O4 and HN2O4 (26-fluoro-4.7.13.16-tetraoxa-1,10-diaza-tri- cyclo[8.8.7.1,20,24]hexacosa-20,22,24(26)-triene and 4,7,13,16-tetraoxa-1,10-diaza-tricyclo[8.8.7.1,20,24]hexacosa-20,22,24 (26)-triene, respectively) and FN,O, and HN2O3 (23-fluoro-4,7,20-trioxa-1,10-diazatricvclo[8.7.5.1,12,16]tricosa-12,14, (23)-triene and 4,7,20-trioxa-1,10-diazatricyclo[8.7.5.1,12.16]tricosa-12,14,16( triene, respectively) [optimized synthesis yields for FN,O, of 73% and for FN2O4 of 70%] were investigated by potentiometry and calorimetric titrations. A comparison of the complexes of the fluoro cryptands FN2O3 and FN2O3, with those of the closely related macrocycles HN2O4 and NN,04 provides information on the stabilizing effect of CF-M interactions, since the two types of ligand differ only in the substitution of the fluorine atom by hydrogen in the latter compounds. In most cases the fluoro cryptands form more stable complexes (up to 3.5 logk( units) with the metal ions. This phenomenon is most pronounced when the radius of the metal ion and the size of the macrocyclic cavity are complementary. Mismatched metal complexes of, for example, Cs+ with FN2O4 or HN2O4 are of equal stability, since the metal ion is too large to be coordinated by the CF group within the macrocyclic cavity. The X-ray crystal structure of the most stable complex FN,O,-Ba(C10,), was determined, and its short F-Ba2+ distance of 284.2(2) pm indicates significant stabilization due to F-Ba2+ interactions. Ab initio calculations on the model reactions Li+(OMe2)(3)(C6H5F)-->Li+(OMe2)(3) + C-6. H5F and Li+(C6H5F)-->C6H5F gave Li+-F interaction energies of 43.7 and 78.7kJmol(-1), respectively (counterpoise-corrected MP2 energies on DFT(BP86)-optimized stuctures), These values amount to about 64 and 51%, respectively, of the corresponding calculated Li+(OMe2)(3)- OMe2 and Li+- OMe2 binding energies.