Autism spectrum disorder affects more than 5 million children in south Asia. Although early interventions have been used for the treatment of children in high-income countries, no substantive trials have been done of the interventions adapted for use in low-income and middle-income countries (LMICs). We therefore assessed the feasibility and acceptability of the parent-mediated intervention for autism spectrum disorder in south Asia (PASS) in India and Pakistan. A single-blind randomised trial of the comparison of 12 sessions of PASS (plus treatment as usual) with treatment as usual alone delivered by non-specialist health workers was done at two centres in Goa, India, and Rawalpindi, Pakistan. Children aged 2-9 years with autism spectrum disorder were randomly assigned (1:1) by use of probabilistic minimisation, controlling for treatment centre (Goa or Rawalpindi), age (<6 years or ≥6 years), and functional impairment (Vineland Adaptive Behaviour Scale Composite score <65 or ≥65). The primary outcome was quality of parent-child interaction on the Dyadic Communication Measure for Autism at 8 months. Analysis was by intention to treat. The study is registered with ISRCTN, number ISRCTN79675498. From Jan 1 to July 30, 2013, 65 children were randomly allocated, 32 to the PASS group (15 in Goa and 17 in Rawalpindi) and 33 to the treatment-as-usual group (15 in Goa and 18 in Rawalpindi). 26 (81%) of 32 participants completed the intervention. After adjustment for minimisation factors and baseline outcome, the primary outcome showed a treatment effect in favour of PASS in parental synchrony (adjusted mean difference 0·25 95% CI 0·14 to 0·36; effect size 1·61 95% CI 0·90 to 2·32) and initiation of communication by the child with the parent (0·15 0·04 to 0·26; effect size 0·99 0·29 to 1·68), but time in mutual shared attention was reduced (-0·16 -0·26 to -0·05; effect size -0·70 -1·16 to -0·23). Our results show the feasibility of adapting and task-shifting an intervention used in a high-income context to LMICs. The findings also replicate the positive primary outcome treatment effects of a parent-mediated communication-focused intervention in the original UK Preschool Autism Communication Trial, with one negative effect not reported previously. Autism Speaks, USA.