Franson's Bell experiment with energy-time entanglement [Phys. Rev. Lett. 62, 2205(1989 does not rule out all local hidden variable models. This defect can be exploited to compromise the security of Bell inequality-based quantum cryptography. We introduce a novel Bell experiment using genuine energy-time entanglement, based on a novel interferometer, which rules out all local hidden variable models. The scheme is feasible with actual technology.PACS numbers: 03.65. Ud, 03.65.Ta, 03.67.Mn, 42.50.Xa Two particles exhibit "energy-time entanglement" when they are emitted at the same time in an energyconserving process and the essential uncertainty in the time of emission makes undistinguishable two alternative paths that the particles can take. Despite this fundamental deficiency, and despite that this defect can be exploited to create a Trojan horse attack in Bell inequality-based quantum cryptography [6], Franson-type experiments have been extensively used for Bell tests and Bell inequality-based quantum cryptography [7], have become standard in quantum optics [8,9], and an extended belief is that "the results of experiments with the Franson experiment violate Bell's inequalities" [9]. This is particularly surprising, given that recent research has emphasized the fundamental role of a (loophole-free) violation of the Bell inequalities in proving the device-independent security of key distribution protocols [10], and in detecting entanglement [11].Polarization entanglement can be transformed into energy-time entanglement [12]. However, to our knowledge, there is no single experiment showing a violation of the Bell-CHSH inequality using genuine energy-time entanglement (or "time-bin entanglement" [13]) that cannot be simulated by a LHV model. By "genuine" we mean not obtained by transforming a previous form of entanglement, but created because the essential uncertainty in the time of emission makes two alternative paths undistinguishable.Because of the above reasons, a single experiment using energy-time entanglement able to rule out all possible LHV models is of particular interest. The aim of this Letter is to describe such an experiment by means of a novel interferometric scheme. The main purpose of the new scheme is not to compete with existing interferometers used for quantum communication in terms of practical usability, but to fix a fundamental defect common to all of them.We will first describe the Franson Bell-CHSH experiment. Then, we will introduce a LHV model reproducing any conceivable violation of the Bell-CHSH inequality. The model underlines why a Franson-type experiment does not and cannot be used to violate local realism. Then, we will introduce a new two-photon energy-time Bell-CHSH experiment that avoids these problems and can be used for a conclusive Bell test.The Franson Bell-CHSH experiment.-The setup of a Franson Bell-CHSH experiment is in Fig. 1. The source emits two photons, photon 1 to the left and photon 2 to the right. Each of them is fed into an unbalanced interferometer. BS i are be...