<p indent="0mm">Recently, the future of particle physics and high energy accelerators aroused extensive attention from society. The view that current accelerators will soon be replaced by those based on the new acceleration mechanism, is widely spread. This view implies that next-generation high energy accelerators shall not be built until the new mechanism is ready. This paper introduces the near and mid-term science goals of particle physics, and the needed accelerators with the energy and luminosity requirements. By analyzing current status and possible future achievements of the new acceleration mechanism, or more precisely, plasma wakefield acceleration mechanism driven by laser or particle beams, we conclude that traditional accelerators will still be the main choice in the next <sc>20–30 years.</sc> First of all, we argue that studies such as neutrino physics, cosmic-rays physics, dark matter searches and dark energy studies may not rely on accelerators, but they do not represent the main-stream of particle physics. After the discovery of the Higgs boson, the particle physics community reached a consensus that our next priority is the study of Higgs couplings with all the other elementary particles. This study will not only test precisely the Standard Model, but also be the best probe to new physics beyond the Standard Model. For this purpose, a very high luminosity Higgs factory, based on the electron-positron collision, is needed. After a short introduction of different types of traditional accelerators and their related technologies, we conclude that linear electron (positron) accelerators are limited by their acceleration capabilities while their luminosity is much smaller than circular electron (positron) accelerators. Circular electron (positron) accelerators are limited by the synchrotron radiation hence their energy cannot reach <sc>300 GeV</sc> or above for a reasonable power consumption. Below this energy limit, circular colliders have advantages on luminosity, power consumption, cost, etc., over linear ones. Circular proton accelerators have very little synchrotron radiation, hence the acceleration goal is easy to achieve but the bending power of the magnet becomes the limit of energy. For linear proton accelerators, they have no advantages at all on energy, luminosity, power consumption, cost, etc., except at very low energies. From the above analysis, it is evident that the plasma-acceleration mechanism can only play a role on linear accelerators to improve acceleration capabilities if luminosity is not highly demanding. Obviously, even if plasma-accelerators are applicable, they are still not suitable for Higgs factories where the luminosity is a key while the energy is in the comfort zone of circular accelerators. Laser-driven accelerators are still far away to be applicable since their average power, power efficiency, beam quality and beam intensity are several orders of magnitude off, while beam-driven accelerators are off by a few factors to one order of magnitude. In addition, issues such as positron and proton acceleration mechanism, high efficiency multi-stage cascade acceleration need to be resolved and experimentally demonstrated. It is a common understanding that there are still <sc>20–30 years</sc> for the new acceleration mechanism to be a viable choice of high energy accelerators. However, plasma-based accelerators may find applications in irradiation, radiation imaging, free-electron laser, etc. in coming years. It may also be combined with traditional accelerators to form an injector. The future of high energy accelerators becomes then very clear: We shall build a circular electron-positron collider as a Higgs factory while actively working on R&D for possible future very high energy (1–10 TeV) e<sup>+</sup>e<sup>–</sup> linear colliders, as well as on high <italic>T</italic><sub>c</sub> superconducting technologies, in particular iron-based superconducting materials for future circular proton colliders.