Background: Complete fusion cross sections in collisions of light, weakly bound nuclei and high Z targets show suppression of complete fusion at above-barrier energies. This has been interpreted as resulting from breakup of the weakly bound nucleus prior to reaching the fusion barrier, reducing the probability of complete charge capture. Below-barrier studies of reactions of 9 Be have found that breakup of 8 Be formed by neutron stripping dominates over direct breakup, and that transfer triggered breakup may account for the observed suppression of complete fusion.Purpose: This paper investigates how the above conclusions are affected by lifetimes of the resonant states that are populated prior to breakup. If the mean life of a populated resonance (above the breakup threshold) is much longer than the fusion timescale, then its breakup (decay) cannot suppress complete fusion. For shortlived resonances, the situation is more complex. This work explicitly includes the mean life of the short-lived 2 + resonance in 8 Be in classical dynamical model calculations to determine its effect on energy and angular correlations of the breakup fragments and on model predictions of suppression of cross sections for complete fusion at above-barrier energies.Method: Previously performed coincidence measurements of breakup fragments produced in reactions of 9 Be with 144 Sm, 168 Er, 186 W, 196 Pt, 208 Pb and 209 Bi at energies below the barrier have been re-analysed using an improved efficiency determination of the BALiN detector array. Predictions of breakup observables and of complete and incomplete fusion at energies above the fusion barrier are then made using the classical dynamical simulation code PLATYPUS, modified to include the effect of lifetimes of resonant states.Results: The agreement of the breakup observables is much improved when lifetime effects are included explicitly. Sensitivity to sub-zeptosecond lifetime is observed. The predicted suppression of complete fusion due to breakup is nearly independent of Z, and has an average value of ∼ 9%. This is below the experimentally determined fusion suppression which is typically ∼ 30% in these systems.Conclusions: Inclusion of resonance lifetimes is essential to correctly reproduce breakup observables. This results in a larger fraction of nuclei remaining intact at the fusion barrier radius, compared with calculations that do not explicitly include lifetime effects. The more realistic treatment of breakup followed in this work leads to the conclusion that the suppression of complete fusion cannot be fully explained by breakup prior to reaching the fusion barrier. Only one third of the observed fusion suppression can be attributed to the competing process of breakup. Other mechanisms that can suppress complete fusion must therefore be investigated. One of the possible candidates in cluster transfer that produces the same heavy target-like nuclei as those formed by incomplete fusion. PACS numbers: 25.40.Hs,25.70.Hi,25.70.Pq,25.70.Ef It was recognised early on [11] that ver...