Telepsychology research has focused primarily on treatment efficacy, with far less attention devoted to how common factors relate to teletherapy outcomes. This research identified trajectories of depressive symptom relief in 105 older people living with HIV with elevated depressive symptoms enrolled in a randomized clinical trial testing two 12‐session group teletherapies and compared common factors (e.g., therapeutic alliance and group cohesion) across depressive symptom trajectory groups. Growth mixture modelling of weekly depression scores identified three depressive symptom change groups: (1) ‘early improvers’ (31%) who reported reductions in depressive symptoms by Session 4; (2) ‘delayed improvers’ (16%) whose symptoms improved after Session 5 and (3) ‘non‐improvers’ (53%). Therapeutic alliance was unrelated to treatment outcome group. Group cohesion was greater in early improvers than non‐improvers. Group cohesion was unexpectedly lower, and group member similarity was greater in delayed improvers than non‐improvers. Early improvers had been living with HIV/AIDS for fewer years than non‐improvers. In group teletherapy, group cohesion and group member similarity are more important than client–therapist alliance. Copyright © 2015 John Wiley & Sons, Ltd. Key Practitioner Message In group teletherapy with older people living with HIV (OPLWHIV), three latent outcome trajectory groups emerged over the 12‐week treatment period: (1) non‐improvers (53%); (2) early improvers (31%) and (3) delayed improvers (16%). In group teletherapy with OPLWHIV, group cohesion is a stronger predictor of depressive symptom relief than is client–therapist alliance. OPLWHIV in group teletherapy who do not respond to treatment until the latter therapy sessions can still experience depressive symptom relief comparable with early responders.