The compositions of 54 sherds (ca. 1751 to the early 1900s) from three Worcester porcelain factory sites were determined by electron microprobe. The data show that the early proprietors of the first (Warmstry House) factory initially (1751) produced a mildly phosphatic frit (lead-bearing) ware. Soapstone was added to this paste (and flint glass frit virtually eliminated) after Worcester acquired the Lund's Bristol works early in 1752. It is not known when in the early or mid-1750s they abandoned this transitional paste in favor of their long used soapstone-frit paste. One transitional phosphatic sherd has an unusually high magnesium content (16.5 wt.% MgO), indicating the use of ∼48 wt.% soapstone, probably added to improve the paste's plasticity. A nonphosphatic, experimental ware is unusually lead-rich (17.1 wt.% PbO), apparently reflecting an attempt to achieve a more vitrified ware than the earlier phosphatic porcelain. After about the mid-1750s, relatively minor variations were made to the Mg-Pb paste, but some wares produced during the Flight and Barr period (1792-1804) have compositions intermediate between true (Si+Al-rich) porcelain and lead-free soapstone wares. True porcelain sherds were recovered from Robert Chamberlain's Diglis factory (1783-1840), which later became Chamberlain & Co. and, subsequently, was operated by Kerr and Binns who produced bone china. At the St. Martins Gate works, the Graingers manufactured true porcelain and bone china. Both Grainger wares used a novel, variably lead-bearing, barium-rich glaze. In 1862, the above-mentioned factories amalgamated under the auspices of Royal Worcester, which, like Grainger, produced true porcelain and bone china but, instead, used barium-free, alkali-lime, and lead-bearing glazes on these two types of ware, respectively. The geochemical data indicate that the history of porcelain manufacture in Worcester is one of both continuity and innovation. [ABSTRACT FROM AUTHOR]