Introduction: Due to the restrictions imposed to control the COVID-19 pandemic, there has been an increase in studies based on online surveys. However, there are important concerns about the validity and generalizability of results from online surveys. Thus, we aimed to test the reliability and validity of the online version of the International Physical Activity Questionnaire short form (IPAQ-SF) among college students from low-income regions. Methods: This was a methodological feasibility study with a random stratified sample from a college located in the state of Maranhão in the city of Imperatriz (Brazil). The sample consisted of 195 college students (at least 17 years of age) to evaluate the validity and 117 students to evaluate the reliability. All data were collected in a self-reported online format (via Google Forms) twice, with an interval of 2 weeks. We used Spearman’s correlation analysis for the reliability study. Additionally, we applied exploratory and confirmatory factor analysis to evaluate the structural validity. Results: The questionnaire showed acceptable (rho > 0.30) and significant (p < 0.05) reliability, except for the question about the duration of sitting time on a weekend day. When assessing the construct validity (exploratory analysis), we identified a single factor that explained 88.8% of the variance. The 1-factor model showed acceptable model fit (SRMR = 0.039; CFI = 0.96; TLI = 0.90) in confirmatory analysis. Conclusions: The online version of the IPAQ-SF has acceptable reliability among college students from low-income regions and maintains the structure of the construct regarding to physical activity.