Objective To test the effectiveness of a brief self-compassion intervention in improving mental health outcomes for mothers of infants. Method A randomized controlled trial study design was used. A community sample of mothers of infants (<2 years) completed measures of self-compassion, fears of compassion, psychological flexibility, depression, anxiety, stress, symptoms of posttraumatic stress, and infant feeding experiences. Mothers randomized to intervention received access to online self-compassion resources, and 248 mothers (interventionn = 94, waitlist-controln = 154) completed postintervention assessment 8 weeks later. Results Overall, 62.8% (n = 59) of intervention participants accessed the resources per-protocol, and lower fear of compassion scores predicted resource use. At postintervention, mothers who used the resources had improved scores for posttraumatic stress symptoms (95% confidence interval [CI] = 0.31-5.47,p = .028), depression (95% CI = 0.15-2.01,p = .023), self-compassionate action (95% CI = 0.41-3.45,p = .012), and engagement with compassion from others (95% CI = 0.22-5.49,p = .034) compared to waitlist-control. Fears of compassion moderated intervention effectiveness. There were no effects on other outcome variables. Conclusions Findings support the potential effectiveness of interventions based on compassion-focused therapy to improve maternal mental health.