The potential for new 4-, 5-, and 6-junction solar cell architectures to reach 50% efficiency is highly leveraging for the economics of concentrator photovoltaic (CPV) systems.The theoretical performance of such next-generation cells, and experimental results for 3- and 4-junction CPV cells, are examined here to evaluate their impact for real-world solar electricity generation. Semiconductor device physics equations are formulated in terms of the band gap-voltage offset Woc (Eg/q) − Voc, to give a clearer physical understanding and more general analysis of the multiple subcell band gaps in multijunction cells. Band gap-voltage offset is shown experimentally to be largely independent of band gap Eg for a wide range of metamorphic and lattice-matched semiconductors from 0.67 to 2.1 eV. Its theoretical Eg dependence is calculated from that of the radiative recombination coefficient, and at a more fundamental level using the Shockley-Queisser detailed balance model, bearing out experimental observations. Energy production of 4-, 5-, and 6-junction CPV cells, calculated for changing air mass and spectrum over the course of the day, is found to be significantly greater than for conventional 3-junction cells. The spectral sensitivity of these next-generation cell designs is fairly low, and is outweighed by their higher efficiency. Lattice-matched GaInP/GaInAs/Ge cells have reached an independently confirmed efficiency of 41.6%, the highest efficiency yet demonstrated for any type of solar cell. Light I-V measurements of this record 41.6% cell, of next-generation upright metamorphic 3-junction cells with 40% target production efficiency, and of experimental 4-junction CPV cells are presented. Copyright © 2010 John Wiley & Sons, Ltd.