The Early Start Denver Model is an evidence-based, comprehensive naturalistic developmental behavioral intervention for young and very young children with autism spectrum disorder. This interdisciplinary model is suitable for the Taiwanese public health system. In this study, we evaluated the effectiveness of an Early Start Denver Model program with adaptation, including a lower intensity, shorter duration, and delivery in general hospitals by multidisciplinary professionals. A quasi-experimental study with pre-, post-intervention, and 6-month follow-up was conducted. Forty-five young children with autism spectrum disorder, aged 2-4 years, were grouped into the Early Start Denver Model and control (the usual community treatment) groups. The Early Start Denver Model group received 6 months of one-on-one intervention for approximately 8-9 h per week. Compared with the control group, the Early Start Denver Model group showed greater improvements in overall development quotient and nonverbal development quotient from pre- to post-intervention, but these differences (a Group-by-Time interaction) did not sustain at the 6-month follow-up. Being mindful of some potential threats to internal validity associated with clinical trial designs and implementation, this study provides preliminary evidence to support the effectiveness of the Early Start Denver Model intervention in regional general hospital settings in the context of Han-Chinese-mainly culture.