Abstract This study was carried out with the principal aim of obtaining reliable outcomes for the future implementation of a temperature-phased anaerobic digestion (TPAD) process in a large (2 M population equivalent, p.e.) WWTP. With the aid of pilot-scale (10 L) reactors fed by pure primary sludge (PS), a TPAD process, where the first and the second reactor were operated at 50 °C and 38 °C, respectively, was compared with a conventional mesophilic (38 °C) anaerobic digestion (AD) process. The initial hydraulic retention time (HRT) of the first, acidogenic, reactor of the TPAD was reduced from 3 to 2 days in the second part of the test. The results demonstrated that the TPAD system had been stable for all the duration of the test (approx. 100 days), as testified by the steady values of pH and tVFAs/TA ratio, notwithstanding the decrease in the HRT. The TPAD proved to be more efficient in volatile solid (VS) reduction and methane generation, compared to the conventional mesophilic AD process. In fact, the VS reduction increased from 42% to approx. 55% and the specific methane potential (SMP) from 280 to 332 NL/kg VS added. An excellent phase separation was observed between the two acidogenic and methanogenic reactors, as demonstrated by the low SMP (only 3% of the overall production) recorded from the first reactor of the TPAD system. However, the energy analysis demonstrated that the higher SMP obtained in the TPAD was not sufficient to compensate the higher amounts of heat required for sludge heating and heat loss compensation. Only a process of heat recovery could make the TPAD system really profitable, thus increasing the aliquot of energy in the form of methane, available for users external to the WWTP, by 20%. This result represents a step in the evolution of traditional WWTPs towards more energy efficient and sustainable facilities.