Several studies have reported a link between poor adolescent sleep and academic achievement. Evidence suggests that particular intellectual abilities that are integral in academic achievement may be susceptible to sleep loss. One such ability is working memory. The aim of the present study was to investigate the link between adolescent sleep loss and working memory performance. A total of 143 adolescents (aged 13–18 years, mean age = 14.9 ± 1.4 years, 43% boys) volunteered for the study. All participants completed an online sleep questionnaire and the working memory tasks of letter–number sequencing and operation span task. Participants were grouped into sufficient sleepers (i.e. more than 9 h; n = 43), borderline sleepers (8 to 9 h; n = 57) and insufficient sleepers (less than 8 h sleep; n = 43), based on their self-reported total sleep time on school nights. The adolescents reporting insufficient sleep performed worse on both letter–number sequencing and operation span task (both P = 0.03; medium to large effect sizes) compared with borderline sleepers, but not compared with sufficient sleepers. No other significant differences between groups were found. Compared to the other groups, insufficient sleepers reported going to bed later (P < 0.0001), taking longer to fall asleep (P < 0.0001) and experiencing greater daytime sleepiness (P < 0.05). These findings suggest between 8 to 9 h of sleep on school nights is sufficient for optimum working memory performance. Future experimental studies are needed to support the current study's findings.