SummaryA total synthesis of the antitumor sesquiterpene coriolin (9; racemic) in 11 steps from 3,3,6-trimethylbicyclo[2.2.2]oct-7-ene-2,.5-dione (2a/2b) is described (yield 2a/ 2b-t8: 28 "/o). The sequence is unprecedentedly short and avoids difficult separation problems. The key step in the scheme is a novel facet of oxadi-n-methane photochemistry, i.e., the steering by subtle steric effects of the P,y-unsaturated E-diketone to undergo a regioselective photorearrangement involving one P,y -enone partial chromophore. Furthermore, the overall phototransformation, which can be carried out at unusually high concentrations ( > 20% solutions), involves also a Norrish type 1 process equilibrating the two epimeric starting enediones 2a and 2b in favour of the desired stereoisomer.The antitumor sesquiterpene coriolin, (-)-(9)'), has become a testing case for the design of syntheses of medium-sized natural products of complex structures, and notably for the development of new methodologies aiming at cyclopentanoid annulation procedures [4]. We have previously demonstrated that the use of the optically active forms of tricyc1o[3.3.0.0*~*]octan-3-one provides for a versatile synthetic approach to cyclopentanoid natural products [3] [5] [6], including one towards the total synthesis of coriolin, (-)-(9) [6]. Scheme 1 summarizes an alternative to this concept which is characterized by an unprecedentedly short route to the intermediate 8 in common with previous total syntheses of (f)-coriolin (9) [2]. The key step in this scheme is a novel facet of oxadi-n-methane photochemistry, i e . , the steering by subtle steric effects of a P,y -unsaturated E -diketone to undergo a regioselective photorearrangement involving one P,y-enone partial chromophore.') Synthetic work directed towards ( * )-9, see [I]; synthesis of 8 and its conversion into (%)-9, see [4. For the isolation of (-)-9 and its biological importance, see Footnote 82 in [3].