The (+)-cis-chromanochromanones (2) and (3), representing the basic ringsystem of the natural insecticide rotenone, have been synthesised by a general procedure; the key step is 4-aroylation of the 4-phenyl-sulphonyl)chromans (7; Z=S02Ph). Rotenone (1) is the principle insecticide of Derris resin, a natural preparation at one time widely employed in agriculture; it is still utilised as a horticultural and domestic pesticide, and as a piscicide in fish farming. Rotenone blocks mitochondrial electron transport at Complex I in insects, but is biodegradable and is rapidly detoxified in animals, minimising environmental health hazards. 1 This combination of properties has made rotenone an important target for structure-activity studies; however such work has concentrated on compounds derived from the natural series, because of the difficulties associated with total synthesis. To overcome this limitation on the development of the rotenoid group as modern pesticides, new approaches to the chromanochromanone ring system are required, combining brevity with accessible starting materials. We have addressed this problem, and this paper sets out one new strategy, leading to synthesis of the parent rotenoid (2), (formally (+)-6a,l2a-dihydro-6Hrotoxen-12-one)and of its relative (3). Such work is relevant to areas other than agrochemistry, since rotenoids also display antimicrobial and antiviral activity, and are potent inhibitors of SRS-A. We adopted the strategy set out retrosynthetically in the Scheme. Opening of ring C(4) in the retro-Michael sense would lead to a I-aroylchrom-3-en, further disconnectible to a vinyl anion or its synthetic equivalent. This appeared an attractive prospect since (a) natural rotenoids are known to epimerise at the H/C junction in mild base, with retention of cis-stereochemistry, 2 showing that cis-(4) is thermodynamically preferred to (5) i and (b) the vinyl anion equivalent could derive from chroman-4-ones, available in variety. This synthon required an anion stabilising group 2, amenable to elimination as HZ from the putative intermediate (6). Our initial targets were thus chromans (7), and we chose to investigate the series with R1 and R2=H or OMe and with Z=SPh, SOPh, and S02Ph. We expected that the corresponding chromanones (8) would be easy to access by the well known route3 from a suitable phenol, via condensation with a O-halopropionate, hydrolysis to an aryloxypropionic acid (9), and cyclisation in acid. These methods work, but the first step is surprisingly low yielding, and we resorted to the addition of phenols to acrylonitrile,