Knowledge tracing aims to predict students’ future performance based on their historical interaction records. In the research domain of addressing this problem using deep learning methods, Monitoring Student Progress for Learning Process-Consistent Knowledge Tracing (LPKT-S) not only focuses on improving the accuracy of predicting student performance but also strives to simulate a more realistic consistency with the learning process. LPKT-S is currently the model with the best performance in this field. However, it does not explore or utilize factors related to question and concept difficulty that can influence students’ knowledge states. This paper introduces, for the first time, both knowledge concept difficulty and question difficulty into LPKT-S, proposing a Learning Process Consistency and Difficulty-aware Knowledge Tracing Model (LPDKT). The proposed model incorporates question difficulty and knowledge concept difficulty into question representation and integrates these features in the learning gate to better characterize the relationship between students’ knowledge states and the difficulty levels of questions they practice. Additionally, the model introduces knowledge concept difficulty into the forgetting gate to more accurately simulate the decay of students’ knowledge states over time, calculating their progress in continuous learning interactions through the positive impact of learning gains and the negative impact of learning forgetting. Finally, comparative experiments on three public datasets show that LPDKT achieves improvements in AUC and ACC ranging from 1.3% to 14.2% and 1% to 7.7%, respectively, demonstrating its superior predictive performance.