Purpose <p>Patients with food insecurity and chronic diseases that require long durations of infusion treatment (e.g., cancer) are more likely to delay or forego treatment and have worse outcomes than food-secure patients. Our aim was to test feasibility, adherence, and exploratory effects of providing different forms of assistance on patient-reported food security, diet, and well-being.</p> Methods <p>Adults receiving treatment at an infusion clinic in Texas were screened for food insecurity. Eligible patients were referred to a mixed-methods pilot RCT and received an assistance service for 3 months: pantry only referrals, pantry + 12 medically tailored meals (MTM) per month, or pantry + $75 USD payments.</p> Results <p>Patients (<i>N</i> = 150) were screened from January to April 2024; 54 were eligible and 45 enrolled. Thirty-nine completed (87%), two dropped out (4%), and four entered hospice (9%). Only 54% of completers used the pantry, with most only using it once (<i>N</i> = 12) or twice (<i>N</i> = 5). Of those receiving MTM, 100% received it for 1 month, 80% for 2 months, and 67% for all months. There were statistically significant improvements in perceived diet quality over time, but no interaction by group. There were increases in fruit and vegetable consumption among those receiving MTM and non-significant decreases in other groups. Well-being increased among all but was not significant. Food security improved in pantry only and MTM but slightly decreased in the payment group.</p> Conclusion <p>The study was feasible, but adherence to monthly pantry appointments was low. Future studies could explore whether there is greater adherence with co-located pantries and greater benefit from longer-term support or larger quantities of food and/or payments.</p>