Diesel particulate filter (DPF), as part of aftertreatment system of internal combustion engine, is considered to be the only feasible way to prominently lessen particle emissions under the requirement of today’s strict regulations such as Euro Ⅵ, US Tier 3 and China Ⅵ. This paper gives a brief introduction of the mechanism and regeneration approaches of DPF, with emphasis on soot load estimation inside the filters, which plays a vital role in formulating regeneration control strategy and ensuring exhaust systemic dependability. Various methods are covered according to different principles, including differential-pressure based methods, which are mostly used nowadays, novel model-based methods and also several newfangled soot sensors, which are progressively developed to meet the increasingly stringent on-board diagnosis (OBD) requirements. The focus of future soot detection and quantitative prediction is to improve accuracy, reliability and robustness, which may necessitate consideration of soot distribution, ash effect, failure identification and fault tolerance handling.