Background: Substance use among college students remains an important public health issue, which may have been exacerbated by the social and behavioral health disturbances resulting from the COVID-19 pandemic. Objectives: This investigation employed egocentric social network analysis to assess associations between individual and network-level factors on students' self-reported changes in substance use behaviors 18 months into the pandemic. Methods: In Fall 2021, an online Qualtrics survey was administered at a large public university in the southwestern United States. Enrolled college students (n = 355) completed the survey voluntarily. An egocentric network analysis explored how personal network metrics influence substance use at the individual level. Multinomial logistic regression models assessed individual and network-level factors, relative to changes in students' substance use behavior patterns. Three distinct models were computed, predicting changes in (1) alcohol consumption, (2) smoking behavior, and (3) recreational drug use. Results: The majority of respondents reported no change to their smoking/vaping (54%) and recreational drug use (43%); however, alcohol use increased for a large portion of respondents (41%). When just comparing the overall proportion of respondents who either decreased or increased their substance use during the pandemic, more were found to have increased consumption of alcohol, tobacco, or other drugs as opposed to consuming less. For instance, more than half of respondents (56%) indicated their alcohol consumption had increased, while 4 out of every ten respondents indicated their recreational drug use increased. Respondents who increased their substance use during the prior 18 months perceived greater substance use behaviors among peers in their networks. Conclusions: Future investigations should explore unique network properties and influences on substance use behaviors and changes among college students in higher education institutions.