
We calculate the nexttoleadingorder corrections to the perturbative QCD calculation of the decay J/psi arrow pi+pi that cannot be factorized into either initial or final state wave functions. When using either the asymptotic or ChernyakZhitnitsky distribution amplitudes for the pion we find that these corrections are small (approximately 5%). Since the leading process is proportional to F squared subscript pi (M 2sj), we suggest that the pion form factor at a large timelike Q2 can be extracted from the measured decay rate of charmonium. Normalizing to the corresponding leptonic decay channels, we thus obtain that M2J F pi (M J) is nearly equal to 0.9, approximately a factor of 2 larger than what might have been commonly expected.

