We present observational evidence that eruptions of quiescent filaments and associated coronal mass ejections (CMEs) occur as a consequence of the destabilization of large-scale coronal arcades due to interactions between these structures and new and growing active regions. Both statistical and case studies have been carried out. In a case study of a ''bugle'' observed by the High-Altitude Observatory Solar Maximum Mission coronagraph, the high-resolution magnetograms from the Big Bear Solar Observatory show newly emerging and rapidly changing flux in the magnetic fields that apparently underlie the bulge. For other case studies and in the statistical work the eruption of major quiescent filaments was taken as a proxy for CME eruption. We have found that two thirds of the quiescent-filament-associated CMEs occurred after substantial amounts of new magnetic flux emerged in the vicinity of the filament. In addition, in a study of all major quiescent filaments and active regions appearing in a 2-month period we found that 17 of the 22 filaments that were associated with new active regions erupted and 26 of the 31 filaments that were not associated with new flux did not erupt. In all cases in which the new flux was oriented favorably for reconnection with the preexisting large-scale coronal arcades; the filament was observed to erupt. The appearance of the new flux in the form of new active regions begins a few days before the eruption and typically is still occurring at the time of the eruption. A CME initiation scenario taking account of these observational results is proposed. ¿ American Geophysical Union 1995