U ovom radu analizirano je buduće stanje klime za 24 lokacije na području Hrvatske: Zagreb, Krapina, Sisak, Karlovac, Varaždin, Koprivnica, Bjelovar, Rijeka, Gospić, Virovitica, Požega, Slavonski Brod, Zadar, Osijek, Šibenik, Vukovar, Split, Pazin, Dubrovnik, Čakovec, Blato, Knin, Sinj i Zavižan. Napravljena je analiza srednjih, maksimalnih i minimalnih dnevnih temperatura te klimatskih indeksa: broj ljetnih dana, broj uzastopnih ljetnih dana, trajanje toplinskog vala u odnosu na srednje trajanje dobiveno za referentno razdoblje i broj toplih dana određen prema 90. percentilu u referentnom razdoblju. U analizi su korištene simulacije dva regionalna klimatska modela RegCM4 i CLMcom-CCLM4-8-17 horizontalne rezolucije 12.5 km. Regionalni modeli forsirani su s 3 različita globalna klimatska modela, EC-EARTH, MPI-ESM i HadGEM2-ES, uz uvažavanje dva moguća scenarija buduće klime, RCP4.5 i RCP8.5. Temperature i indeksi su analizirani za razdoblje od 1971. do 2070. Modelirane srednje temperature u povijesnom razdoblju (1971. – 2000.) su uspoređene s odgovarajućim srednjacima istog niza dobivenog iz mjerenih podataka na postajama DHMZ-a. Uočen je značajan porast srednjih, maksimalnih i minimalnih temperatura u svim korištenim kombinacijama regionalnih i globalnih klimatskih modela za sve promatrane lokacije, a veći porast dobiven je za RCP8.5 nego za RCP4.5. Iako se čini da je dobivena stopa porasta temperatura generalno veća za priobalne lokacije nego za kontinentalne, odnosno za južnije više nego za sjevernije, RegCM4 model naginje predviđanju viših stopa porasta temperatura za kontinentalne lokacije, npr. za Čakovec i Koprivnicu. RegCM4 model predviđa da će Čakovec, VaraŽdin i Koprivnica imati najveću stopu porasta srednjih i maksimalnih, a Bjelovar, Koprivnica i Pazin imati najveće stope porasta minimalnih dnevnih temperatura. CLMcom-CCLM4-8-17 pokazuje da će najveću stopu porasta srednjih i maksimalnih temperatura imati Rijeka i Split za RCP8.5 odnosno Rijeka i Pazin za RCP4.5, a najveću stopu porasta minimalne temperature imati Karlovac i Zagreb za RCP8.5 odnosno Pazin i Rijeka za RCP4.5. Na svim se promatranim lokacijama vidi značajan linearni rast svih indeksa. Veći broj dana s maksimalnom temperaturom zraka iznad 25◦C, u svim kombinacijama modela, uočava se na maritimnim postajama, a najveći je u Blatu i Zadru. Slični rezultati dobiveni su i za broj uzastopnih dana s maksimalnom temperaturom zraka iznad 25◦C gdje su pak u prosjeku Blato, Zadar i Šibenik imali najveći porast. Najveći porast toplih dana s obzirom na 90. percentil referentnog razdoblja imaju Blato i Zadar, dok najveći porast trajanja toplinskog vala u odnosu na srednju vrijednost referentnog razdoblja (ref. razdoblje 1971. – 2000.) imaju Split i Sinj za model CLMcom-CCLM4-8-17, a Gospić i Karlovac za RegCM4. Iako je analiza provedena za 24 lokacije diljem Hrvatske, radi sažetijeg prikaza u ovome radu detaljnije se prikazuju samo rezultati za Zagreb i Dubrovnik. In this paper, the future state of the climate was analyzed for 24 locations in Croatia: Zagreb, Krapina, Sisak, Karlovac, Varaždin, Koprivnica, Bjelovar, Rijeka, Gospić, Virovitica, Požega, Slavonski Brod, Zadar, Osijek, Šibenik, Vukovar, Split, Pazin, Dubrovnik, Čakovec, Blato, Knin, Sinj and Zavižan. Mean, maximum and minimum daily temperatures were analysed together with climate indices: number of summer days, number of consecutive summer days, heat wave duration compared to the average duration obtained for the reference period and number of hot days determined according to the 90th percentile in the reference period. Simulations of two regional climate models RegCM4 and CLMcom-CCLM4-8-17 with a horizontal resolution of 12.5 km were used in the analysis. The regional models were forced with 3 different global climate models, EC-EARTH, MPI-ESM and HadGEM2-ES, considering two possible future climate scenarios, RCP4.5 and RCP8.5. Temperatures and indices were analyzed for the period from 1971. to 2070. Mean temperatures in the historical period of modeled data (1971. – 2000.) were compared with corresponding averages of the same series obtained from measured data at DHMZ stations. A signiffcant increase in mean, maximum and minimum temperatures was observed in all considered combinations of regional and global climate models for all observed locations, and a greater increase was observed for RCP8.5 compared to RCP4.5. Although it seems that the obtained rate of temperature increase is generally higher for coastal locations than continental ones, i.e. more for southern than for northern ones, the RegCM4 model tends to predict higher rates of temperature increase for continental locations, e.g. for Čakovec and Koprivnica. The RegCM4 model predicts that Čakovec, Varaždin and Koprivnica will have the highest rate of increase in mean and maximum, and Bjelovar, Koprivnica and Pazin will have the highest rates of increase in minimum daily temperatures. The CLMcom-CCLM4-8-17 model shows that Rijeka and Split will have the highest rate of increase in mean and maximum temperatures for RCP8.5 and Rijeka and Pazin for RCP4.5. Karlovac and Zagreb will have highest rate of minimum temperature increase for RCP8.5 and Pazin and Rijeka for RCP4.5. In all observed locations, a signiffcant linear growth of all indices can be seen. A greater number of days with a maximum air temperature above 25◦C, in all model combinations, is observed at maritime stations, and the highest one is in Blato and Zadar. Similar results were obtained for the number of consecutive days with a maximum air temperature above 25◦C, where, on average, Blato, Zadar and Šibenik had the largest increase. Blato and Zadar have the largest increase in warm days with regard to the 90th percentile of the reference period, while Split and Sinj have the largest increase in the duration of the heat wave in relation to the mean value of the reference period (ref. period 1971. – 2000.) for the CLMcom-CCLM4-8-17 model, and Gospić and Karlovac for RegCM4. Although the analysis was conducted for 24 locations throughout Croatia, for the sake of a more concise presentation, only the results for Zagreb and Dubrovnik are presented in more detail.