Objectives <p>There are limited interventions for preventing excessive infant weight gain among Hispanic and low-income populations. Here, we describe the development/implementation of the Baby-Act trial, a multi-component/multi-model intervention for preventing excessive infant weight gain in low-income Hispanics. We also report adherence/fidelity, lessons learned, and future applications.</p> Methods <p>We conducted a cluster-randomized controlled trial among mother/infant participants of the WIC program in Puerto Rico. The intervention included responsive feeding, infant activation, healthy sleep, and limiting sedentary behaviors using behavioral change theories. It was implemented as a multi-modal approach from pregnancy to the first year of life through weekly online interactive lessons (<i>n</i> = 56) that focused on one of the topics, reinforced with weekly text messages and monthly calls with a dietitian. Adherence/fidelity was recorded as the number of lessons/calls completed and text messages sent/received in the order designed.</p> Results <p>Adherence from 207 participants was 45.7% for pregnancy lessons and 31.4% for baby lessons (higher in months 0–3 and for lessons that suggested uploading images/videos). On average, 19 lessons were completed (with 3.6 lessons completed on the same day instead of once per week). Adherence to calls was 42.3% (higher in the first months, 49.0%). On average, each participant completed 5 calls. Lessons/calls/texts were completed/received as designed; 99.3% of text messages were sent.</p> Conclusions <p>Adherence to all intervention components was low, but all participants received the main messages because the topics were repeated continuously. Future studies should reduce the frequency of lessons, include more text messages, and more sharing of videos/photos.</p> Clinicaltrials.gov <p>NCT03517891.</p>