Hi Heiko, > Scaling BABAR neutral B to charged B results in (1.15+-0.33)10^-4. > Belle measured for B+ -> rho0lnu: (1.39+-0.29)10^-4. I see, for rho0 l nu the difference between BaBar and Belle is a little larger than for rho-l nu. Anyway, the errors for the two analyses are not too far apart (except that BaBar limited by syst., Belle limited by stat.), so using only Belle would not really help and I would still suggest to use BaBar for collab-internal consistency. If you're determined to beat down the 3% uncertainty by combining the results, we should maybe come up with an average of our own in the next few days. Please let me know. Thanks, Jochen