Background: Treatment burden is the workload of being a patient and its impact on wellbeing. Little is known about change in treatment burden over time for people with multimorbidity. Aim: To quantify change in treatment burden, determine factors associated with this change, and evaluate a revised single-item measure for high treatment burden in older adults with multimorbidity. Design and setting: A 2.5-year follow-up of a cross-sectional postal survey via six general practices in Dorset, England. Method: GP practices identified participants of the baseline survey. Data on treatment burden (measured using the Multimorbidity Treatment Burden Questionnaire; MTBQ), sociodemographics, clinical variables, health literacy and financial resource were collected. Change in treatment burden was described, and associations assessed using regression models. Diagnostic test performance metrics evaluated the single-item measure relative to the MTBQ. Results: 301 participants were recruited (77.6% response rate). Overall, there was a 2.6% increase in treatment burden. 98 (32.6%) and 53 (17.6%) participants experienced an increase and decrease, respectively, in treatment burden category. An increase in treatment burden was associated with having more than five long-term conditions (ß:8.26 (95% CI: 4.20 to 12.32) and living >10 minutes (vs.≤10 minutes) from the GP (aß:3.88 (95% CI: 1.32 to 6.43)), particularly for participants with limited health literacy (mean difference: aß:9.59 (95% CI: 2.17 to 17.00)). The single-item measure performed moderately- sensitivity: 56.5%; specificity: 92.5%. Conclusion: Treatment burden changes over time. Improving access to primary care, particularly for those living further away from services, and enhancing health literacy, may mitigate increases in burden.