Background: Children’s conduct and emotional problems increased during the COVID-19 pandemic. Objective: We tested whether a smartphone parenting support app, Parent Positive, developed specifically for this purpose, reversed these effects in a cost-effective way. Parent Positive includes 3 zones. Parenting Boosters (zone 1) provided content adapted from standard face-to-face parent training programs to tackle 8 specific challenges identified by parents and parenting experts as particularly relevant for parents during the pandemic. The Parenting Exchange (zone 2) was a parent-to-parent and parent-to-expert communication forum. Parenting Resources (zone 3) provided access to existing high-quality web-based resources on a range of additional topics of value to parents (eg, neurodevelopmental problems, diet, and sleep). Methods: Supporting Parents And Kids Through Lockdown Experiences (SPARKLE), a randomized controlled trial, was embedded in the UK-wide COVID-19: Supporting Parents, Adolescents and Children during Epidemics (Co-SPACE) longitudinal study on families’ mental health during the pandemic. Parents of children aged 4 to 10 years were randomized 1:1 to Parent Positive or follow-up as usual (FAU) between May 19, 2021, and July 26, 2021. Parent Positive provided advice on common parenting challenges and evidence-based web-based resources and facilitated parent-to-parent and expert-to-parent support. Child conduct and emotional problems and family well-being were measured before randomization (T1) and at 1 (T2) and 2 (T3) months after randomization. Service use, costs, and adverse events were measured, along with app use and satisfaction. The primary outcome was T2 parent-reported child conduct problems, which were analyzed using linear mixed regression models. Results: A total of 320 participants were randomized to Parent Positive, and 326 were randomized to FAU. The primary outcome analysis included 79.3% (512/646) of the participants (dropout: 84/320, 26% on Parent Positive and 50/326, 15% on FAU). There were no statistically significant intervention effects on conduct problems at either T2 (standardized effect=−0.01) or T3 (secondary outcome; standardized effect=−0.09) and no moderation by baseline conduct problems. Significant intervention-related reductions in emotional problems were observed at T2 and T3 (secondary outcomes; standardized effect=−0.13 in both cases). Parent Positive, relative to FAU, was associated with more parental worries at T3 (standardized effect=0.14). Few intervention-attributable adverse events were reported. Parent Positive was cost-effective once 4 outliers with extremely high health care costs were excluded. Conclusions: Parent Positive reduced child emotional problems and was cost-effective compared with FAU once outliers were removed. Although small when considered against targeted therapeutic interventions, the size of these effects was in line with trials of nontargeted universal mental health interventions. This highlights the public health potential of Parent Positive if implemented at the community level. Nevertheless, caution is required before making such an interpretation, and the findings need to be replicated in large-scale, whole-community studies.