This study's purpose was to modify the Screening Tool for Autism in Two-Year-Olds (STAT) into a Taiwanese version called T-STAT. Study 1 included 15 children with Autism and 15 children with Developmental Delay (DD) or language impairment (LI) aged between 24 and 35 months. Study 2 had 77 young children with Autism, PDD-NOS, or DD/LI as a clinical-based validation sample. In Study 1, the signal detection procedure found that a cutoff score of 2 would yield high sensitivity and specificity in T-STAT. In Study 2, using a score of 2 as a cutoff, the agreement between T-STAT risk and ADOS classification was highly acceptable. Results were promising as a Level 2 screening tool for Autism for ages two to three.