Thick loess deposits on the Chinese Loess Plateau provide an outstanding archive of paleoclimate, geomagnetic field and paleoenvironmental changes over at least the last 2.6 Ma. Much work on magnetic polarity reversals, pedogenic magnetic enhancement and climatic change has been carried out over the past 30 years. However, questions about how Chinese loess becomes magnetized remain unanswered. In this thesis, I have sought to address key questions concerning magnetization processes in the Chinese loess. To understand magnetization processes, it is important to understand which magnetic minerals are present and the processes that caused them to be present. Magnetic and other analyses of both loess and paleosol samples indicate that pedogenic superparamagnetic and single domain magnetite and hematite can be present. The mass concentration of hematite can exceed 90% of the magnetic particle assemblage. An eolian dust deposition simulator designed for laboratory redeposition experiments was then used to demonstrate that both a depositional remanent magnetization (DRM) and a post-depositional remanent magnetization (PDRM) can be acquired by Chinese loess. Water content significantly affects the efficiency of PDRM acquisition. No significant PDRM was observed when the water content was below a critical threshold of 40%. Published paleomagnetic data from the Chinese loess indicate the widespread occurrence of inclination error. Detailed paleomagnetic studies of U-channel samples from the last loess/paleosol sequence at Yuanbao were made to assess whether the Chinese loess can accurately record geomagnetic relative paleointensity variations. Three normalized remanence records were determined using conventional normalization as well as the pseudo-Thellier method. Strong coherence among the normalized remanence records indicate that the whole range of magnetic minerals has been affected by pedogenesis. Poor similarity among published normalized remanence records from the Chinese loess, as well as with marine relative paleointensity records indicate that the Chinese loess is not suitable for relative paleointensity determinations. The range of magnetic minerals present and their varying distribution within loess and paleosol sediments indicate that the magnetization is due to a mixture of remanence acquisition mechanisms including DRM, PDRM, chemical remanent magnetization (CRM) and viscous remanent magnetizations, which also indicates complexities in remanence acquisition that will complicate attempts to extract relative paleointensity information. This thesis has provided an improved understanding of the mechanisms of Chinese loess magnetization, and re-emphasizes the importance of pedogenic CRM acquisition.