This research proposes a real-time implementable robust eco-cooperative adaptive cruise control (Eco-CACC) strategy with the consideration of gear shift. Vehicle acceleration is optimized in real-time to vary the vehicle power demand to improve the fuel efficiency. The effects of different gear ratios on the fuel consumption rate are explicitly considered during the optimization process. A robust optimization method is adopted to ensure the controller performance under traffic prediction uncertainties. Using this approach, the optimal solution is guaranteed to provide fuel savings and satisfy constraints at the same time for any actual traffic profile within the uncertainty set. The optimal control problem is discretized and solved using a nonlinear programming (NLP) solver for an 8-vehicle platoon scenario. The results show that the proposed controller can achieve 23.5% fuel saving with perfect traffic prediction, and 16.6% fuel saving with prediction uncertainties.