To investigate the associations between 21 metals in fine particulate matter (PM2.5) and their corresponding concentrations in urine in the general population. Between April and May 2011, this panel study enrolled 120 residents using random sampling approach in Wuhan communities which contained 3 035 subjects. Participants were aged 18 to 80 years and had lived in the sampling buildings for at least 5 years. Data from basic questionnaires, physical examinations, and morning blood and urine samples under fasting conditions were collected. Participants with missing data were excluded. Finally, 83 particpants included. Participants were instructed to use personal air samplers to continuously monitor PM2.5 for 24 h. The following 21 metals were measured in PM2.5 and urine by inductively coupled plasma mass spectrometry: aluminum, titanium, vanadium, chromium, manganese, cobalt, nickel, copper, zinc, arsenic, selenium, rubidium, strontium, molybdenum, cadmium, tin, antimony, barium, tungsten, thallium and lead. The associations between PM2.5 metals and urinary metals were investigated using generalized linear regression models. The age of the study population was (51.5±6.3)years. After adjusting for age, sex, smoking status, BMI, education and income, elevated urinary chromium was significantly associated with increased chromium concentrations in personal PM2.5. The least square means (standard deviation) of urinary chromium in participants classified as having low exposure (<12.491 ng/m(3)), intermediate exposure (12.491-32.388 ng/m(3)) and high exposure (>32.388 ng/m(3)) were (-1.334±0.756), (-1.114±0.813) and (-0.718±0.645) μg/mmol creatinine, respectively (P=0.009). However, the association between urinary and personal PM chromium was not observed after additionally adjusting for false discovery rate (P>0.05). Furthermore, the results demonstrated that other metals in PM2.5 were not related to their corresponding concentrations in urine of subjects. Urinary concentrations of metals did not reflect human exposure to metals in air, and may not be appropriate as an index to assess personal exposure to metals in particulate matter.