Polish com. (Lewis weak acid centers, 285 mmols/g), Aldrich's 504C acidic (191), 507C neutral (208), and 5016A basic (179) Al2O3 (I) were Impregnated (I) with aq. (NH4)H2PO4, dried (2 h, 383 K) and calcined In air (6 h, 773 K) to yield supports contg. 1, 2, 5, and 10% P2O5 (P1...P10), or (ii) with aq. (NH4)(2)MoO4 and H2O evaporating at room temp., dried (4 h, 383 K), and (A) heated in air (4 h, 523 K, 800 Pa) to yield the Mo A series; or (B) calcined In dried air (4 h, 823 K, 1 atm) to yield the MoB series of catalysts. The P2O5-modified I gave the Mo(A)(or B)P1...P10 catalysts. The catalytic test used to determine acidic centers in I Involved dehydration of cyclohexanol to c-C,H,, followed by isomerization and disproportionation to c-C6H14, Me-c-pentane, and C6H6. Catalyst activity was examined In C2H4 metathesis (10 or 20 mL/min over 0.5 g catalyst preactivated in situ at 873 K for 2 h In Ar) at 303 K (1 atm). MoA and MoB annealed In Ar (2 h, 873 K) had BET surfaces of 196 and 204 m(2)/g, resp. Compared with MoO3(10%)I (2.2 Mo atoms/nm(2)), the MoBP1...P10 (and also MoAP1...P10) were the less active the higher was the P content. In MoO3 (3 or 5%)/I, 2 and 5% P2O5 raised the catalyst activity 2 and 4 times, resp. With I replaced by SiO2.Al2O3 I as support In MoO3(10%)/I, the catalyst activity fell; thermal treatment A was superior to B. Strongly acidic OH groups on the support surface disfavor the formation of stable Mo active sites.