Background and Aim: We studied sleep patterns and glucose homeostasis in 215 subjects (156 F; median age 56 years, range 23-70) at risk of type 2 diabetes (T2D) based on the FINDRISC questionnaire enrolled in the PRE-MED2 project (funded by the Tuscany Region). Methods: Subjects underwent a 75-g OGTT for glucose, insulin, and C-peptide. Objective and subjective sleep measures were assessed using the actigraphy, the Pittsburgh Sleep Quality Index score (PSQI), the Epworth Sleepiness Scale (EES) and the STOP-BANG questionnaire. Results: Normal glucose tolerance (NGT) occurred in 94 subjects (43.7%), impaired fasting glucose and/or impaired glucose tolerance (IFG/IGT) in 111 (51.6%), and T2D in 10 (4.7%). BMI (NGT 27.2±5.2; IFG/IGT 29.5±5.5; T2D 35.5±6.2 kg/m2; p<0.001), waist circumference (NGT 93.6±12.1; IFG/IGT 100.5±13.3; T2D 111.2±8.0 cm; p<0.001) and HOMA-IR (NGT 1.44±1.19; IFG/IGT 2.34±1.57; T2D 4.08±1.03; p<0.01) were significantly different among groups. While PSQI and ESS scores were similar, the STOP-BANG score differed among groups (NGT 2.3±1.4; IFG/IGT 2.5±1.4; T2D 3.6±1.1; p=0.006). Total sleep time (TST) (NGT 399.8±107.3; IFG/IGT 383.5±86.9; T2D 342.1±152.4 min; p=0.04) and sleep efficiency (SE) (NGT 86.2±11.7; IFG/IGT 86.4±7.0; T2D 78.1±11.0%; p=0.03) were progressively declined across categories with no differences in wake after sleep onset (WASO) or in sleep regularity index (SRI). Within the whole population, BMI was associated with higher STOP-BANG scores (r=0.36; p<0.01), lower SE (r=-0.24; p<0.001), shorter TST (r=-0.14; p<0.04) and longer WASO (r=0.14; p=0.02). Subjects with high HOMA-IR were more likely to have higher STOP-BANG scores (r=0.26; p<0.001) and higher neck circumference (r=0.39; p<0.01). HOMA-IR was negatively related with SE (r=-0.16; p=0.02) and positively related with WASO (r=0.18; p=0.01). Glycemia at different time points during OGTT was related with STOP-BANG scores (baseline r=0.19, p=0.007; +60’ r=0.24, p=0.001; +120’ r=0.16, p=0.02), but not with sleep-related actigraphic data. EES scores were negatively associated with SRI (r=-0.19; p=0.006). PSQI scores were positively related with WASO (r=0.20; p=0.004) and negatively with SE (r=-0.17; p=0.013). BMI (F-value 6.69; p=0.002) and HOMA-IR (F-value 4.31; p=0.02) were predictors of poor SE and long WASO, respectively. Conclusion: Sleep disorders are more common in people at risk of T2D. BMI and HOMA-IR are predictor of sleep disorders among the whole population.