Diurnal preference (chronotype) is a useful instrument for studying circadian biology in humans. It harbours trait-like dimensions relating to circadian period and sleep homeostasis, but also has ontogenetic components (morningness increases with age). We used the Morningness-Eveningness questionnaire (MEQ) in the Baependi study, a family-based cohort study based in a small town in Minas Gerais, Brazil. The population is highly admixed and has a cohesive and conservative lifestyle. 825 individuals (497 female) aged 18-89 years (average 6 SD 5 46.4 6 16.3) and belonging to 112 different families participated in this study. The average MEQ score was 63.5 6 11.2 with a significant (P , 0.0001) linear increase with age. Morningness was significantly (P , 0.0001) higher in the rural (70.2 6 9.8) than in the municipal zone (62.6 6 11.1), and was also significantly (P 5 0.025) higher in male (64.6 6 10.9) than in female (62.8 6 11.2) participants. Thus, in spite of universal access to electricity, the Baependi population was strongly shifted towards morningness, particularly in the rural zone. Heritability of MEQ score was 0.48 when adjusted for sex and age, or 0.38 when adjusted for sex, age, and residential zone. The reported MEQ score heritability is more akin to those of previous twin studies than previous family studies.C ircadian period is viewed as a stable and reproducible quantitative trait in humans 1 . It has been found to have a high degree of heritability in all vertebrates where it has been studied 2-4 , and super-short periods have been shown to co-segregate with single-gene mutations both in humans 5 and in model animals 6 . Whilst circadian period may be determined experimentally in humans, most feasibly through the forced desynchrony protocol 7 , it is costly, intrusive, and time-consuming, and thus not a feasible option for large-scale human phenotyping. A more practical proxy for such studies is diurnal preference or chronotype, a self-reported questionnaire-based instrument that yields a numerical score on a scale ranging between extreme morning preference and extreme evening preference. It has been shown to correlate significantly not only with circadian phase, but also with circadian period 8 . However, the trait-like dimension of diurnal preference/chronotype is more complex, as in some individuals, morning versus evening preference results from differences in the build-up or dissipation of homeostatic sleep pressure rather than in circadian parameters [9][10][11] .In addition to its endophenotypic qualities, chronotype is also influenced by state-like variables, some of them of considerable interest to our understanding of the interaction between the circadian oscillator and the sleep homeostat and our external environment 12 . There is a firmly established relationship between chronotype and age, with a peak in eveningness in late adolescence 13 followed by a gradual increase in morningness 14 reflecting an altered relationship between the circadian system and sleep-wake timing 8 .The Morningn...