Purpose: Prior work developed a shortened 16-item version of the Social Responsiveness Scale (SRS), a quantitative measure of social communication and autism spectrum disorder (ASD)-related traits. However, its properties for use in risk factor estimation have not been fully tested compared to the full SRS. We compared the associations between gestational age (previously established risk factor for ASD) and the 65-item "full" and 16-item "short" versions of the SRS to test the shortened version's ability to capture associations in epidemiologic analyses of ASD risk factors. Methods: We used data from participants in the Environmental influences on Child Health Outcomes (ECHO) Program (n = 2,760). SRS scores were collected via maternal/caregiver report when children were aged 2.5-18 years. We compared estimates of associations between gestational age and preterm birth between the full and short SRS using multivariable linear regression, quantile regression, and prediction methods. Results: Overall, associations based on full and short SRS scores were highly comparable. For example, we observed positive associations between preterm birth with both full ([beta] =2.8; 95% CI [1.7, 4.0]) and short ([beta] =2.9; 95% CI [1.6, 4.3]) SRS scores. Quantile regression analyses indicated similar direction and magnitude of associations across the distribution of SRS scores between gestational age with both short and full SRS scores. Conclusion: The comparability in estimates obtained for full and short SRS scores with an "established" ASD risk factor suggests ability of the shortened SRS in assessing associations with potential ASD-related risk factors and has implications for large-scale research studies seeking to reduce participant burden.