In 1950 Sheehan and Bose2 reported a convenient synthesis of 4,4-dicarboxy-2-azetidinone p by the cyclization of d-haloamidomalonates 1. The intermediate 1 is readily available by the acylation of aminomalonates with a suitable u-halo acid but since this method cannot be extended to the preparation of a-amid0 or d-alkoxy B-lactams (&,R' = NHCOR" or OR"), we have devised an alternative approach which we describe here. Condensation of ketomalonic acid esters proceeds smoothly with primary amines to give the imine ,$ in high yield. In recent years we have used the reaction of a variety of acid chlorides with imines in presence of triethylamine to prepara B-lactams a-substituted' with func-Q tional groups such as N3, OR, OCR, Br, etc. This acid chloride-triethylamine reaction proved equally applicable to the imine $; thus, phenoxyacetyl chloride, triethylamine and $ gave the desired B-lactam 2. The ester groups of 2 can be utilized for modifying the B-lactam: mild saponification of the diester .6-lactam 2 produced the crystalline mono ester B-lactam 8 of "El' configuration because saponification of the carboethoxy group which is truns to the phenoxy group in 2 (and therefore less hindered) can be expected to proceed faster than that of the cis ester group. Decarboxylation of 6 in refluxing pyridine generated the cis-d-lactam [ (J3"_kH=5 Hz) suggesting thereby that decarboxylation had proceeded with retention of configuration. Since no epimerization could be expected under these mild conditions, the method reported here provides a facile pathway for the stereoselective synthesis of cis-3,4-disubstituted-2-azetidinones4. Treatment of ,!$ with two equivalents of sodium hydroxide in dioxane led to 4,4-dicarboxy azetidinone ,Q in 35% yield, a versatile intermediate in the synthesis of N-substituted aspartic acid derivatives 2,s . c @,a-Dicarboxy-a-lactam ,Q was also utilized for the synthesis of the Spiro-B-lactam barbiturate ,$ through condensation with diisopropyl carbodiimide in 17% yield. Substituted barbituric acids are of course of considerable current interest in medicinal chemistry. In the light of our earlier work it can be expected that various analogs of 2 will 3547