The sleep-wake cycle of adolescents is characterized by a phase delay associated with biological, behavioral, and social changes, which are affected by school schedules. The aim of this study was to analyze the relationship between school shifts and sleep habits and quality, daytime sleepiness, and cognitive performance in adolescents from two urban contexts. A total of 294 adolescents from technical high school courses participated in the study, aged 15.3 ±0.8 years (175 girls/119 boys). A total of 133 went to classes in the morning shift and 161 in the afternoon shift. They were subdivided into two groups according to the degree of urbanization: less urbanized (Region 1) and more urbanized (Region 2). Sleep pattern was assessed by the following questionnaires: “Health and Sleep”, Pittsburgh Sleep Quality Index, Puberty and Phase Preference Scale, Pediatric Daytime Sleepiness Scale, in addition to a 10-day sleep diary. In these 10 days, cognitive tasks were performed to assess attention and working memory. Morning shift students showed earlier bedtimes (B=-72; p=0.00) and wake up times (B=-177.3; p=0.00) and had less time in bed (B=-103.7; p =0.00) in the week, worse sleep quality (B=1.5; p=0.00), greater irregularities in waking up time (B=43.3; p=0.00) and time in bed ( B=120.1; p=0.00) between weekdays and weekends, and higher rates of social jetlag (B=31.0; p=0.00), greater daytime sleepiness (B=1.5; p=0.04), greater frequency of naps during the week and predominance of school schedule as a reason for waking up comparing to those from the afternoon shift. Although students from the morning shift had a lower percentage of correct answers in selective attention (B=-5.5; p=0.03) and in tonic alertness (B=-2.0; p=0.09), they presented shorter reaction times in phasic alertness (B=-36.8; p=0.03). Regarding the degree of urbanization, students from Region 2 showed later bedtimes (B=-49.7; p=0.00) and wake up times (B=-31; p=0.01), had less time in bed (B=22.8; p=0.06) and greater sleepiness upon waking up (B=-0.7; p=0.00), as well as a greater occurrence of activities that favor alertness at night and longer reaction times in the phasic alertness (B=-46.7; p=0.01). However, they had a higher percentage of correct answers in tonic alertness (B=-4.1; p=0.00), phasic (B=-7.2; p=0.00) and selective attention (B=-10, 1; p=0.00); and lower percentage of omissions in tonic (B=1.5; p=0.04) and phasic (B=1.6; p=0.03) alertness. Overall, the tendency towards evening was associated with later bed (B=-5.4; p=0.00) and wake up times (B=-3.6; p=0.00), higher sleepiness (B =-0.3; p=0.00) and worse sleep quality (B=-0.2; p=0.00), without association with cognitive variables. There was no consistent relationship between the memory variables and the predictor ones. Thus, future studies using other working memory assessment tasks are needed to analyze this relationship. Therefore, our study confirms the hypothesis that the beginning of classes at 7:00 am is associated with impairments in sleep habits and attentional performance of adolescents, being affected also by the degree of urbanization. We suggest that educational opportunities experienced throughout life can affect the attentional performance of adolescents, which can differ between urban contexts.