Severe obesity during childhood is associated with cognitive deficits. Studies in adults have suggested improvements in executive functioning and memory after bariatric surgery. Our aim was to explore changes in cognitive function in adolescents over two years after bariatric surgery or intensive non-surgical treatment. The Adolescent Morbid Obesity Surgery 2 (AMOS2) is a multicentre, open-label, randomised controlled trial in which adolescents (aged 13-16 years) with severe obesity (defined as body mass index (BMI) ≥35kg/m2) at three specialised obesity centres in Sweden, were randomly assigned to receive bariatric surgery or intensive non-surgical treatment. Herein we report the results of the prespecified exploratory endpoint of change in cognitive functioning. Inclusion in AMOS2 required Tanner pubertal stage ≥3, previous participation in lifestyle obesity treatment for at least one year, and passed assessment form a paediatrician and a paediatric psychologist. Adolescents with severe intellectual disability or other severe, pervasive developmental disorder were excluded. Participants underwent baseline assessment of general intellectual ability, executive functioning, and memory before randomisation. Tests were administrated by clinical psychologists and repeated at one- and two-year follow-up timepoints. Differences in means between groups during follow-up are provided with confidence intervals. The trial is registered at ClinicalTrials.gov, NCT02378259. Between October 28 2015 and June 7 2017, 46 adolescents (74% girls), with a mean age of 15.8 (±0.92) years and a mean BMI of 42.8 (±5.4) kg/m2, were included and randomised (23 to bariatric surgery and 23 to intensive non-surgical treatment). At baseline 23/46 (50%) of the adolescents had general intellectual functioning classified as borderline or below. For 15/18 (83%) aspects of cognitive functioning, no significant differences in change over two years were identified between groups; Immediate (average difference during follow-up 1.0 [95% CI:-2.6 to 4.6]) and Delayed (0.5 [95% CI:-0.6 to 1.6]) Verbal Recall, Category Fluency (1.1 [95% CI:-1.6 to 3.8]) and Switching (1.5 [95% CI:-0.0 to 2.9]), Number (-6.0 [95% CI:-12.3 to 0.3]) and Letter (0.1 [95% CI:-5.2 to 5.3]) Sequencing, Number-Letter Switching (-10.3 [95% CI:-26.4 to 5.8]), Motor Speed (-8.3 [95% CI:-17.5 to 0.9]), Colour Naming (-1.9 [95% CI:-4.2 to 0.3]), Inhibition (-3.6 [95% CI:-9.6 to 2.5]), Inhibition Switching (-6.7 [95% CI:-15.3 to 1.9]), Mazes (-0.5 [95% CI:-4.9 to 3.9]), Digit Span Forward (0.1 [95% CI:-0.6 to 0.9 ]) and Backward (0.6 [95% CI:-0.4 to 1.6 ]), and Estimated IQ (0.4 [95% CI:-3.9 to 4.8]; all p>0.05). Three sub-tests assessing fundamental cognitive skills improved more over two years in operated adolescents than in intensive non-surgical treatment; Letter Fluency (average difference during follow-up 3.8 [95% CI: 0.1-7.5]; p=0.046), Visual Scanning (-6.5 [95% CI:-11.6 to-1.5]; p=0.011), and Word Reading (-1.9 [95% CI:-3.3 to-0.4]; p=0.011). In contrast to non-randomised studies in adults, we could not demonstrate an association of bariatric surgery and its accompanying significant weight loss with overall greater improvement in executive functions and memory in adolescents over two years compared with a non-surgical group without weight loss. However, lack of statistical power is a potential limitation. The clinical relevance of greater improvements in basic cognitive skills needs to be explored. Sweden's innovation agency (VINNOVA), Swedish Research Council, Joanna Cocozza foundation for paediatric research, The Skane University Hospital Psychology Research and Development Grant, Tore Nilsson's Foundation, SUS Foundations and Donations, and Mary von Sydow's Foundation.