OBJECTIVE: This study examines the association between social support and cognitive function among midlife and older MSM living with or without HIV. DESIGN: We analyzed longitudinal data from participants enrolled from October 2016 to March 2019 in the Patterns of Healthy Aging Study, a substudy of the Multicenter AIDS Cohort Study. METHODS: We conducted a cross-sectional analysis to estimate the association between social support and three measures of cognitive function [Trail Making Test (TMT) Part A, TMT Part B to A ratio, and Symbol Digit Modalities Tasks (SDMT)]. We also used linear mixed-effects models to estimate the association between baseline social support and cognitive function across four subsequent time points. We evaluated a multiplicative interaction term between baseline social support and time, in order to determine whether cognitive trajectories over time vary by baseline social support. RESULTS: Social support was associated with lower TMT Part A scores at baseline and over the subsequent 2 years, indicating better psychomotor ability. Social support was associated with higher SDMT scores at baseline and across 2 years, indicating better information processing. We observed no association between social support and TMT B to A ratio at baseline or across 2 years, indicating no effect on set-shifting ability. Longitudinal cognition outcome trajectories did not vary by the level of baseline social support. CONCLUSION: Social support and cognitive function were associated in this sample over a short time period. Further research should explore causal relationships over the lifespan.