Obtaining α from B→ρ±π∓

We discuss how to extract the weak phase α from present data on B→ρ±π∓ decays. Introducing αeff and constraining the difference from α using flavor SU(3), one arrives at α=(95±16)∘, if a testable assumption of a small relative strong phase between the two relevant tree amplitudes is employed to distinguish between discrete ambiguities. On the long run we advocate the combined fit to B→ρ±π∓ and SU(3) related modes. The effect of SU(3) breaking in this approach is expected to be small, because of relatively small penguin pollution.