In this article, a novel guidance law derivation and new synchronization strategy are proposed for a virtual structure-based formation flight. These are designed using both aerodynamic and dynamics equations of aerial robots to facilitate implementation of the guidance and control laws. The guidance commands are derived in the form of acceleration based on a new analytical approach. These acceleration commands are converted to suitable inputs for the control system in the form of velocity, roll and pitch angles by employing an innovative strategy. In addition, a new synchronization strategy for virtual structure formation control is proposed. In this strategy, each agent utilizes the other agents’ actual position rather than their position errors. The proposed strategy is capable of shape formation flight using a passive sensor, such as vision sensors, for position detection of neighbour agents. This ability makes the proposed strategy more reliable than conventional synchronization methods. The mentioned strategy is further improved by self-tuning of the synchronization gain based on a fuzzy inference. The simulation of formation flight for a group of three fixed-wing aerial robots using six degrees of freedom models for each one reveals the merits of the proposed strategy. In fact, this approach significantly decreases the number of oscillations and corresponding amplitudes of position/orientation error for each agent. This is a crucial aspect of the mission performance for formation flight control.