<![CDATA[Background/Context: Reading comprehension remains a major educational challenge among Filipino learners, highlighting the need for engaging and contextually relevant instructional approaches. Objective/Purpose: This study examined the effectiveness of Localized Multimedia Instruction developed using Canva AI as an instructional design tool in improving the reading comprehension of Grade 9 learners. Method: A quasi-experimental pretest-posttest control-group design with qualitative support was conducted among 60 Grade 9 learners from two intact classes in a public secondary school in Bulacan, Philippines. The experimental group received four weeks of Localized Multimedia Instruction, while the control group received conventional instruction. Reading comprehension was measured using a validated 40-item test. Independent-samples t-test and paired-samples t-tests were used to examine baseline equivalence and within-group changes, respectively, while ANCOVA was used to compare posttest performance between groups while controlling for pretest scores. Qualitative data were gathered from 28 of the 30 learners in the experimental group through a post-implementation feedback form and analyzed using conventional content analysis. Results: Baseline reading comprehension did not differ significantly between the groups, t(58)=−1.287, p=.203. Both groups demonstrated significant pretest-to-posttest improvement. In the experimental group, the mean score increased from 10.67 (SD = 3.73) to 28.50 (SD = 6.74), representing a mean gain of 17.83 points, 95% CI [16.59, 19.08], t(29)=−29.31, p < .001, Cohen's d = 5.35. After controlling for pretest scores, the experimental group obtained a significantly higher adjusted posttest mean (X̄ = 27.84, 95% CI [26.59, 29.09]) than the control group (X̄ = 14.83, 95% CI [13.58, 16.08]), with an adjusted mean difference of 13.02 points, 95% CI [11.24, 14.80], F(1,57) = 214.47, p < .001, partial η² = .790. Qualitative analysis generated three themes: engaging multimedia learning experiences, contextualized learning, and interactive learning. Conclusion: Localized Multimedia Instruction developed using Canva AI was associated with substantially higher reading comprehension performance and positive learner experiences. The findings suggest that AI-assisted instructional material development can support reading instruction when integrated with localized content, multimedia learning principles, and structured teacher guidance. Findings should be interpreted within the limitations of the quasi-experimental design, two intact classes, absence of individual randomization, and single-school setting.]]>
