The potential health risk caused by long-term exposure to heavy metals in household dust is not only depended on their total content, but also bioaccessibility. In this study, twenty-one dust samples were collected from residential buildings, schools, and laboratories in 14 provincial-capital/industrial cities of China, aiming to evaluate the total contents, fractionation, bioaccessibility and health risks of nine heavy metals (As, Cd, Cr, Ni, Pb, Mn, Zn, Fe, and Cu). Results showed that the highest levels of Cd, Cr, Ni and Zn were found in laboratory dust, As, Pb and Mn in school dust, and Fe and Cu in residential dust, indicating different source profiles of the heavy metals. The mean bioaccessibility of the heavy metals across all samples as evaluated using SBRC (Solubility Bioavailability Research Consortium), IVG (In Vitro Gastrointestinal), and PBET (Physiologically Based Extraction Test) assays was 58.4%, 32.4% and 17.2% in gastric phase (GP), and 24.9%, 21.9% and 9.39% in intestinal phase (IP), respectively. Cadmium had the highest content in the fractions of E1+C2 (43.7%), as determined by sequential extraction, and Pb, Mn, and Zn had a higher content in E1+C2+F3 (64.2%, 67.2%, 78.8%), resulting in a higher bioaccessibility of these heavy metals than others. Moreover, the bioaccessibility of most heavy metals was inversely related to dust pH (R = −0.18 in GP; −0.18 in IP; P < 0.01) and particle size, while a positive correlation was observed with total organic carbon (R = 0.40 in GP; 0.38 in IP; P < 0.01). The exposure risk calculated by the highest bioaccessibility was generally lower than that calculated by the total content. However, Pb in one school dust sample had an unacceptable carcinogenic risk (adult risk = 1.19 × 10−4; child risk = 1.08 × 10−4). This study suggests that bioaccessibility of heavy metals in household dust is likely related to geochemical fractions and physical/chemical properties. Further research is needed to explore the sources of bioaccessible heavy metals in household dust.