The recent development and use of autism measures for the general population has led to a growing body of evidence which suggests that autistic traits are distributed along a continuum. However, as most existing autism measures were designed for use in children older than age 4, to date, little is known about the autistic continuum in children younger than age 4. As autistic symptoms are evident in the first few years, to address this research gap, the current study tested the preschool version of the Social Responsiveness Scale (SRS‐P) in children aged 2 to 4½ years in clinical (N = 74, average age 40 months, 26–51 months) and community settings (N = 357, average age 39 months, 25–50 months) in Japan. Using information obtained from different raters (mothers, other caregivers, and teachers) it was found that the scale demonstrated a good degree of internal consistency, inter‐rater reliability and test‐retest reliability, and a satisfactory degree of convergent validity for the clinical sample when compared with scores from diagnostic “gold standard” autism measures. Receiver operating characteristic analyses and the group comparisons also showed that the SRS‐P total score discriminated well between children with autism spectrum disorder (ASD) and those without ASD. Importantly, this scale could identify autistic symptoms or traits distributed continually across the child population at this age irrespective of the presence of an ASD diagnosis. These findings suggest that the SRS‐P might be a sensitive instrument for case identification including subthreshold ASD, as well as a potentially useful research tool for exploring ASD endophenotypes. Autism Res 2017, 10: 852–865. © 2017 International Society for Autism Research, Wiley Periodicals, Inc.