With earlier initiation and more effective antiretroviral therapy regimens, people with HIV (PWH) are achieving longer lifespans but with high rates of chronic disease and physical function impairment. Regular exercise can reduce disease burden, but older adults with HIV may experience unique barriers to exercise. Exercise self-efficacy (ESE) is an important component of initiation and continuation of exercise, and ESE measurement may illuminate barriers to exercise. Our goal was to evaluate differences in ESE between older PWH and seronegative participants (controls) prior to exercise initiation and over 24-weeks of a supervised resistance and endurance exercise intervention. Virally suppressed PWH and controls completed 12 weeks of moderate-intensity exercise, followed by randomization to 12 weeks of moderate- or high-intensity exercise. The Self-Efficacy for Exercise survey was completed at weeks 0, 12, and 24. Enrolled participants (PWH = 32, controls = 37) had a median age of 56 years [interquartile range: 52, 62]; 63 (91%) were male, and 23 (33%) identified as ethnic or racial minorities. ESE scores differed by serostatus prior to exercise initiation (PWH: 61.3 [95% confidence interval: 54.5, 68]; controls: 73.8 [67.6, 80.1]; p = 0.01; t = - 2.7). ESE scores did not change over the initial 12 weeks in PWH (4.8 [- 2.1, 11.8]; p = 0.17; t = 1.4) or controls (- 2.3 [- 8.9, 4.4]; p = 0.50; t = - 0.7) or between weeks 13 and 24 (pooled serostatus: - 4.8 [- 9.7, 0.2]; p = 0.06: t = - 1.9). Future interventions should incorporate additional strategies to improve ESE, such as text messaging support, coaching, or external rewards.