[en] Recent recoil polarization measurements in Hall A at Jefferson Lab show that the ratio of the electric to magnetic form factors for the proton decreases signi cantly with increasing Q2. This contradicts previous global analyses of Rosenbluth measurements which indicate approximate scaling of the form factors (Gp E(Q2)/Gp M(Q2) H 1). I will present a reanalysis of the cross section data designed to determine if the disagreement between the Rosenbluth extraction and the recoil polarization measurements can be explained by problems with speci c data sets or by changes in the cross-experiment normalization factors.