Epidemiological studies provide evidence of the harmful effects of source-specific fine particulate matter (PM_2.5) on human health. Studies regarding relative contributions of multiple sources to personal exposure are limited and inconsistent. Personal exposure monitoring for PM_2.5 was conducted in 48 adult subjects (ages 18‒63 years) in Hong Kong between June 2014 and March 2015. We identified seven sources of personal PM_2.5 exposure using Positive Matrix Factorization (PMF). These sources included regional pollution (associated with coal combustion and biomass burning), secondary sulfate, tailpipe exhaust, secondary nitrate, crustal/road dust, and shipping emission sources. For personal PM_2.5 exposure, one additional source related to individuals' activities was found: non-tailpipe pollution (characterized by Fe, Mn, Cr, Cu, Sr). We also applied principal component analysis (PCA) for PM_2.5 source identification. The results revealed similar factor/component profiles using PMF and PCA, with some discrepancies in the number of factors. PCA/absolute principal component scores (PCA/APCs) coupled with a linear mixed-effects model (LMM) was applied to the same dataset for source apportionment, adjusting for temperature and relative humidity. Furthermore, stratified PCA/APCs-LMM models were applied to estimate season- and group-specific source contributions of personal PM_2.5 exposure. A mixed source contributions of secondary sulfate, secondary nitrate, and regional pollution were shown (35.1–43.6%), with no seasonal or subject group differences (p > 0.05). Shipping emissions were ubiquitous, contributing 6.3–8.8% of personal PM_2.5 exposure for all subjects. Tailpipe exhaust and traffic-related particles varied by season (p < 0.01) and subject group (p < 0.05). Caution should be taken when using source-specific PM_2.5 as proxies for the corresponding personal exposures in epidemiological studies.