Considering the strong correlation between packet's dissemination range and its originating application's benefit in vehicular networks, this paper introduces 2 new concepts of distance-dependent application influence and time-dependent application accuracy, both of which comprise the application profit. Then, we propose an application-driven multihop broadcast (ADMB) method through hop-by-hop tradeoff between the application profit and network cost in deciding whether or not to relay the broadcast. The dynamic network states and elapsed broadcasting value are joined into the decision-making process, by such way ADMB enables the cross-layer joint optimization between application and network layers. Moreover, ADMB permits applications themselves to preferentially customize their desired dissemination distance and tolerant sensitivity to the time-lag effect, which are characterized by influence coefficient and accuracy coefficient, respectively. Every relaying participant also could customize its priority consideration about application influence and accuracy as well as selfish extent in making decision. We perform extensive numerical calculations and simulations to investigate the applicability and use of ADMB against various combinations of application requirements and network parameters. The results show that ADMB can behave superiorly over the traditional fixed-hop broadcast, fully use the restrictive network resources, and adaptively react to the time-varying network context. KEYWORDS application driven, joint optimization, multihop broadcast, vehicular network 1 Int J Commun Syst. 2017;30:e3306.wileyonlinelibrary.com/journal/dac