Condensation-induced jumping of droplets on superhydrophobic surfaces has received extensive attention because of its great potential for applications in areas such as condensation enhancement and self-cleaning. However, the jumping efficiency of droplets on flat superhydrophobic surfaces is very low, and there is no reliable means of achieving efficient droplet jumping on large scales, which greatly limits its application. To this end, we developed a class of honeycomb bionic superhydrophobic surfaces (HBSS) that enable reliable and efficient droplet jumping on a large scale for the first time and performed experimental and simulation studies on droplet condensation and jumping on this kind of surface. Condensation experiments show that condensate droplets on HBSS can be effectively positioned under the influence of gravity and the uniformity of the droplet diameter is ensured, laying the foundation for achieving efficient jumping. The shape and geometric parameters of HBSS have a significant impact on the droplet jumping efficiency, and the maximum dimensionless jumping velocity of droplet jumping was experimentally measured to be 0.747, corresponding to an efficiency of about 45.25%. Combining with the results of simulation calculations, we found that the surface structure of HBSS can promote more of the excess surface energy to net upward kinetic energy along an extremely efficient and simple pathway (direct conversion), thus achieving an energy conversion efficiency of over 45%.