ImportanceCancer survivors experience accelerated functional decline that threatens independence and quality of life. Previous studies have suggested that vegetable gardening may improve diet, physical activity, and physical function in this vulnerable population, which comprises more than 5% of the US population.ObjectiveTo assess whether diet, physical activity and functioning, and other outcomes improved in older cancer survivors assigned to a vegetable gardening intervention compared with a waitlist.Design, Setting, and ParticipantsFrom May 11, 2016, to May 2, 2022, a 2-arm, assessor-blinded, crossover-designed, intent-to-treat, randomized clinical trial was conducted at cancer survivors’ homes across Alabama. Medicare-eligible survivors of cancers with 5-year survival of 60% or more were registry ascertained and screened for suboptimal vegetable and fruit consumption (<5 servings per day), physical activity (<150 moderate-to-vigorous minutes per week), and physical function (36-Item Short Form Health Survey [SF-36] subscale score ≤90). Consented participants underwent baseline assessments, were randomly assigned to intervention or waitlisted arms, and were reassessed at 1-year follow-up.InterventionOne-year, home-based vegetable gardening intervention providing gardening supplies and mentorship by cooperative extension–certified master gardeners to plant and maintain spring, summer, and fall gardens. Waitlisted participants received the identical intervention after 12 months.Main Outcomes and MeasuresThe main outcome was a composite index of improvements in self-reported vegetable and fruit consumption, physical activity, and physical function corroborated by plasma α-carotene levels, accelerometry, and physical performance assessments, respectively.ResultsOf 381 enrolled participants (mean [SD] age, 69.8 [6.4] years; range, 50-95 years; 263 [69.0%] female), 194 were assigned to the gardening intervention and 187 were waitlisted (attrition rates, 7.2% and 7.0%, respectively). Intent-to-treat analyses did not detect a significant improvement in the composite index of vegetable and fruit intake, moderate-vigorous physical activity, and physical function (intervention arm vs waitlisted arm, 4.5% vs 3.1%; P = .53) or between-arm differences in vegetable and fruit intake (mean difference, 0.3 [95% CI, −0.1 to 0.7] servings per day; P = .10). The intervention arm experienced a significant improvement in vegetable and fruit intake (mean increase, 0.3 [95% CI, 0.0-0.6] servings per day; P = .04). Significant improvements also were observed in the intervention arm vs waitlisted arm in physical performance (mean difference for 2-minute step test, 6.0 [95% CI, 0.8-11.2] steps; P = .03; for 30-second chair stand, 0.8 [95% CI, 0.1-1.5] repetitions; P = .02), perceived health (8.4 [95% CI, 3.0-13.9] points on a 100-point scale [higher scores indicate better health]; P = .003), and gut microbiome alpha diversity (84.1 [95% CI, 20.5-147.6] more observed species; P = .01). The COVID-19 pandemic significantly moderated effects (eg, odds of improvement in self-reported physical functioning were greater before vs during the pandemic: odds ratio, 2.17; 95% CI, 1.12-4.22; P = .02).Conclusions and RelevanceIn this randomized clinical trial including older cancer survivors, a vegetable gardening intervention did not significantly improve a composite index of diet, physical activity, and physical function; however, survivors assigned to the intervention had significantly increased vegetable and fruit consumption and, compared with waitlisted survivors, experienced significant improvements in perceived health and physical performance. Further study in broader populations and during pandemic-free periods is needed to determine definitive benefits.Trial RegistrationClinicalTrials.gov Identifier: NCT02985411