The northeastern Deccan Traps include flows with strong isotopic and elemental affinities to lavas in the southwest as far as 900 km away. However, many of the northeastern flows show a negative correlation of 206Pb/204Pb with ϵNd(t) and positive correlation of 206Pb/204Pb with (87Sr/86Sr)t, distinct from any of the isotopic fields for the southwestern formations; this array (from ϵNd(t)=+5, (87Sr/86Sr)t=0.7041, 206Pb/204Pb=17.9 to ϵNd(t)=0, (87Sr/86Sr)t=0.7068, 206Pb/204Pb=20.5) overlaps the Ambenali Formation (Fm.) field and runs toward that of the common signature <Peng et al., 1994> that appears to have been an important magma type in several of the lower formations of the southwest. This array appears to represent mixing between an Ambenali-type and (1) a common-signature end-member or (2) an end-member with even higher 206Pb/204Pb and 87Sr/86Sr and lower ϵNd than the common signature; such mixing has been postulated, but not documented previously, for the Deccan. Lavas isotopically and chemically indistinguishable from the Ambenali Fm. form the tops of sections near Jabalpur and Chikaldara. Beneath them are flows that chemically resemble the Poladpur Fm. (which sits immediately below the Ambenali in the southwest) and several lavas similar to the Khandala Fm. (beneath the Poladpur and intervening Bushe Fm. in the southwest). Given the general stratigraphic correspondence with the southwestern sequence, many of these northeastern flows may be petrogenetically related to the respective southwestern formations. Lavas with broadly Poladpur- and Khandala-like elemental compositions are also abundant in a thick section south of Mhow; several other Mhow-area flows resemble the distinctive Bushe Fm. However, nearly all of the Poladpur- and Khandala-like northeastern Deccan lavas analyzed have higher 206Pb/204Pb than any southwestern Poladpur or Khandala Fm. basalts. If they indeed are related petrogenetically to these formations, many of the northeastern magmas interacted less with low-206Pb/204Pb continental material and thus must have erupted from different feeder systems. ¿ 1998 American Geophysical Union