In a recent communication [F.J. Vidal-Iglesias, A. Al-Akl, D.J. Watson, G.A. Attard, Electrochem. Commun. 8 (2006) 1147], we demonstrated via a combination of forced deposition and flame-annealing procedures that PtPd {110} single crystal alloy films may be prepared using Pt {110} as a surface template. The present study extends the application of such protocols to the general case of PtPd {hkl}. Characterization of the well-defined PtPd surface alloy films using cyclic voltammetry (CV) and copper underpotential deposition (Cu UPD) allows quantification of the surface alloy composition based on systematic variations in the voltammetric signatures (charge and peak potential) corresponding to Pt-rich and Pd-rich regions of the electrode surface. Comparison with previous Auger/ LEED/CV measurements of alloy composition using bulk PtPd {hkl} electrodes [D.J. Watson, G.A. Attard, Surf. Sci. 515 (2002) 87, T.J. Schmidt, N.M. Markovic, V. Stamenkovic, P.N. Ross Jr., G.A. Attard, D.J. Watson, Langmuir 18 (2002) 6969] provides a number of insights into the structure of the alloy surface and the veracity of the electrochemical estimation of surface composition. For PtPd {111} electrodes, excellent agreement between electrochemical and surface science determinations is observed. This suggests that the original assumption used in the Auger analysis of a homogeneous PtPd film, was correct. However, for the more open PtPd {100} and PtPd {110} surfaces, the disagreement between electrochemical and Auger analyses was more marked. We suggest that in these cases, the assumption of a completely homogeneous selvedge is not correct and strong surface segregation of Pd results in an oscillatory compositional profile as reported for several other Pt binary alloy electrodes. (C) 2007 Elsevier B.V. All rights reserved.