Abstract: ‍ ‍Cell-free massive multiple-input multiple-output （CF-mMIMO） systems randomly deploy numerous distributed access points （APs） within the coverage area to simultaneously serve all users and frequency resources， which can significantly increase the system communication capacity and is one of the most potentially enabling technologies in 6G networks. However， the high power consumption and hardware cost caused by equipping perfect-resolution analog-to-digital converters （ADCs） at numerous APs restrict the practical deployment of CF-mMIMO systems. To effectively reduce the hardware costs， the uplink spectral efficiency （SE） of CF-mMIMO systems with low-resolution ADCs was investigated. Under imperfect channel estimation， a closed-form expression for the user uplink achievable rate in CF-mMIMO systems was derived utilizing the additive quantization noise （AQNM） model and the maximal ratio combining （MRC） receiver filter， and the effect of system parameters such as the number of APs， user transmission power， and ADCs resolution on the SE was analyzed based on this expression. To maximize the SE of the CF-mMIMO system， a greedy pilot assignment algorithm under low-resolution ADCs was proposed to alleviate the pilot contamination. The pilot assignment was modeled as a max-min pilot optimization problem， where the pilot sequence of the user with the lowest rate was iteratively updated to minimize its pilot contamination effect， thus maximizing the achievable rate of this user. Finally， the performance of the CF-mMIMO system with low-resolution ADCs was compared with that of the conventional perfect-resolution ADC system. Numerical simulations reveal that 5-bit low-resolution ADCs nearly match the spectral efficiency （SE） of perfect-resolution ADCs. Furthermore， augmenting the number of antennas at the AP can mitigate the performance loss due to low-resolution ADCs. Moreover， the proposed algorithm not only effectively alleviates the pilot contamination， but also reduces the rate gap among different users and improves the 95%-likely per-user throughput of the system.