Adolescents and young adults (AYAs) with cancer experience substantial symptom burden. Specialty palliative care (SPC) is recommended but often not involved or involved late. To determine whether patient-reported symptom severity was associated with subsequent SPC involvement and whether SPC was associated with symptom improvement in AYAs with cancer. This cohort study comprised AYAs (aged 15-29 years) with primary cancer diagnosed between January 1, 2010, and June 30, 2018, in Ontario, Canada. Data, including self-reported Edmonton Symptom Assessment System (ESAS) scores, were obtained from health care databases. Specialty palliative care was identified through billing codes and validated algorithms. Final data analysis was performed on April 4, 2023. Associations of ESAS scores with subsequent SPC involvement were determined. A difference-in-differences approach was used for patients who died within 5 years of their cancer diagnosis. Case patients (SPC predeath, index date equals first SPC service) were matched 1:1 to control patients (no SPC at equivalent time before death). The study examined whether the difference between 90-day postindex and preindex mean ESAS scores was itself different between case and control patients. This study included 5435 AYAs with cancer, with a median follow-up of 5.1 (IQR, 2.5-7.9) years for analyses of general palliative care. Their median age at cancer diagnosis was 25 (IQR, 22-27) years, and more than half were male (2809 [51.7%]). For all symptoms, moderate and severe ESAS scores were associated with an increased likelihood of SPC involvement compared with mild scores. The greatest magnitude of association was seen for pain scores (hazard ratio for severe vs mild, 7.7 [95% CI, 5.8-10.2]; P < .001). A total of 721 AYAs (13.3%) died within 5 years of diagnosis, and 612 of these patients (84.9%) had received SPC before death. Among 202 case-control pairs, SPC involvement was associated with improved pain trajectories (mean scores improved from 3.41 to 3.07 in case patients and worsened from 1.86 to 2.16 in control patients; P = .003). Other symptom trajectories were not affected. In this cohort study of AYAs with cancer, those reporting moderate or severe symptoms through a screening program were more likely to subsequently receive SPC. These findings suggest that SPC was associated with a subsequent decrease in pain severity but did not affect other symptoms. New interventions targeting other symptoms during treatment and particularly at the end of life are needed.