Objective: Attention control comparisons in trials of stroke rehabilitation require care to minimize the risk of comparison choice bias. We compared the similarities and differences in SLT and social support control interventions for people with aphasia. Data sources: Trial data from the 2016 Cochrane systematic review of SLT for aphasia after stroke Methods: Direct and indirect comparisons between SLT, social support and no therapy controls. We double-data extracted intervention details using the template for intervention description and replication. Standardized mean differences and risk ratios (95% confidence intervals (CIs)) were calculated. Results: Seven trials compared SLT with social support (n = 447). Interventions were matched in format, frequency, intensity, duration and dose. Procedures and materials were often shared across interventions. Social support providers received specialist training and support. Targeted language rehabilitation was only described in therapy interventions. Higher drop-out (P = 0.005, odds ratio (OR) 0.51, 95% CI 0.32–0.81) and non-adherence to social support interventions (P < 0.00001, OR 0.18, 95% CI 0.09–0.37) indicated an imbalance in completion rates increasing the risk of control comparison bias. Conclusion: Distinctions between social support and therapy interventions were eroded. Theoretically based language rehabilitation was the remaining difference in therapy interventions. Social support is an important adjunct to formal language rehabilitation. Therapists should continue to enable those close to the person with aphasia to provide tailored communication support, functional language stimulation and opportunities to apply rehabilitation gains. Systematic group differences in completion rates is a designrelated risk of bias in outcomes observed.