Part I: Attempts were made to synthesise the unknown heterocyclic compound, benz-l-oxepin. Two separate voutes were investigated one involving the use of homochroman-3-one, the other, the use of the isomeric homochroman-5-one, both of which were reducible-to the known homochroman. Homochroman-3-one was prepared by Dieckmann cyclisation of the dimethyl ester of dihydro-o-coumaroxyacetic acid, followed by hydrolysis and decarboxylation of the resultant B-keto-ester. The ketone was converted into the corresponding alphabeta-unsaturated compound which, on spectral evidence, failed to show any enolisation in basic solution. Homochroman-5-one, obtained by cyclisation of g-phenoxybutyric. acid, was treated in a number of ways in an attempt to insert the required double bonds. Work on the olefin, 2:3-dihydro-benz-1-oxepin failed to give the completely unsaturated system required. Both routes were characterised by the large number of by-products obtained at each stage and by the instability of intermediates and these factors almost certainly contributed to the lack of success.