As urban populations grow, the problem of noise pollution becomes more significant. The limited number of epidemiological studies linking individual-level dynamic noise exposure to sleep highlights a gap in our understanding of how individual-level noise exposure impact sleep quality and duration. A cross-sectional survey was conducted in Hong Kong, and portable noise sensors were used to record participants' real-time noise exposure. The Pittsburgh Sleep Quality Index was used to assess their sleep quality. Logistic regression was used to examine the relationship of individual-level noise exposure with sleep quality and sleep duration. Among a total of 763 subjects included in the analysis, and a-weighted decibels [dB(A)] 24-h average exposure (L24) was 63.21 (58.85–67.85) dB(A) for workday and 63.52 (59.03–67.86) dB(A) for non-workday. A significant decrease in individual noise level was observed closer to bedtime (P < 0.05). Compared with the lowest noise exposure, exposure to the highest noise level was positively associated with the presence of poor sleep quality for L24, Ld and Ln, with an odds ratio (OR) and 95 % confidential interval (CI) of 1.53 (1.04–2.24), 1.65 (1.12–2.43), and 1.51 (1.03–2.21), respectively. Exposure to elevated noise level was associated with increased risk of shorter nocturnal sleep duration (< 7 h), the OR and 95 % CI was 1.87 (1.29–2.73) for L24, 1.58 (1.09–2.30) for Ld, 1.50 (1.03–2.18) for Le, and 1.60 (1.10–2.32) for Ln in the highest noise exposure group. Similar findings were observed both on workdays and non-workdays, respectively. Furthermore, the results showed that each 10 dB(A) increment in L24 and Ln was associated with a probable 29 % and 27 % increase in poor sleep quality and a 24 % and 24 % reduction in sleep duration. This study demonstrates that environmental noise exposure was associated with poor sleep quality and shorter sleep duration, posing a significant public health issue in Hong Kong.