Objectives Extended hospitalization for school‐aged cancer patients increases their risk of social marginalization. School‐aged children mature through peer‐interaction, but healthcare providers fail to incorporate this in rehabilitation efforts. The RESPECT study offers classmates to cancer patients to become ambassadors during hospital stays. This study explores classmate decision‐making patterns about ambassadorship. Methods An open‐ended question was prospectively and consecutively provided to classmates (N = 221) (and parents) of 10 children diagnosed with cancer in 2014 and enrolled in the RESPECT study. Statements were analysed using thematic content analysis. Results Of 221 classmates, 140 responded (63%). Of these, 81 applied for ambassadorship (median 8/patient), 58 declined, one was undecided. Nine forms were incomplete; leaving 131 in total that revealed 303 statements for analysis. Five major themes emerged: existing friendship (132/303 statements), personal resources (academic, emotional and social) (107/303), attitudes towards the ambassadorship (34/303), hospital environment (18/303) and logistics (12/303). Of the classmates with pre‐existing friendships, 77% applied for ambassadorship and 80% with a surplus of personal resources applied. These were predominant predictors for ambassadorship application. Classmate motives were condensed into four archetypes: pre‐existing friendship with a surplus of resources (100% applied), non‐friend classmates with a surplus of resources (63% applied), pre‐existing friendship with limited resources (22% applied) and non‐friend classmates with limited resources (0% applied). Conclusion Classmates are highly motivated to support patients during serious illness, irrespective of pre‐existing friendships. Ambassadors offer a novel in‐hospital approach to promote rehabilitation in children with severe/chronic diseases. Results need validation in other settings. Copyright © 2016 John Wiley & Sons, Ltd.