CERES-Wheat model was used to simulate wheat yield with ensemble model data for three time slices (2030–2050, 2050–2070 and 2070–2090) and four representative concentration pathway (RCP) scenarios (2.6, 4.5, 6.0 and 8.5) for four agroclimatic zones (ACZs) of Punjab. The study was conducted for wheat cultivars (HD2967 and PBW725) with two adaptation measures, i.e. shift in sowing dates and shift in sowing date plus additional nitrogen doses. The results showed that 24th November was the optimized sowing date and the yield were higher when supplemented with higher N doses (190 and 230 kg/ha). For the three time slices under RCP2.6 scenario, an increase in yield with combined adaptation measures for four ACZs and respectively for cvs. HD2967 and PBW725 ranged 16–32 and 16–33% (ACZII), 11–39 and 22–43% (ACZIII), 8–47 and 20–51% (ACZIV) and 15–32 and 22–42% (ACZV). Similarly, under stabilization scenarios (RCP4.5 and 6.0) the yield increased with combined adaptation measures and respectively for cvs. HD2967 and PBW725 ranged 15–32 and 14–31% (ACZII), 10–40 and 19–52% (ACZIII), 5–44 and 7–53% (ACZIV) and 14–32 and 20–42% (ACZV). The shortening of the maturity period for the two cultivars ranged 24–34 (ACZII), 21–36 (ACZIII), 9–19 (ACZIV) and 21–32 (ACZV) days under the future scenarios. Under the RCP8.5 scenario, the wheat cultivars were unable to yield higher even with combined adaptation measures during the end of 21st century. Thus, wheat would be a sustainable crop option under climate change in the state, if the sowing date was shifted to 24th November and supplemented with higher N dose.