Autism Spectrum Disorder (ASD) must be present early in development, but may not fully manifest until social demands exceed capacities. In the absence of adequate biological or brain imaging markers to detect and diagnose autism, diagnosis relies on clinical judgment based on observation of symptoms. Many tools have been developed in English-speaking countries (questionnaires for parents, symptom checklists for professionals, observation systems, etc.). Screening in countries with other languages requires cultural and linguistic adaptation of these instruments. This paper presents the adaptation of the ADEC (Autism Detection in Early Childhood). Methods: The original version of the ADEC was translated and culturally and linguistically adapted to the characteristics of the population of Guayaquil (Ecuador). Participants: A pilot study was conducted with a sample of 613 children aged 18-48 months. Results: Cronbach's alpha values (0.89) indicate high internal consistency. The correlation between the MCHAT-R/F follow-up interview and the ADEC (mean r = 0.93) indicates high construct validity. In terms of predictive validity, using the original cut-off points of the ADEC, they show excellent diagnostic ability. The sensitivity and specificity results (sensitivity 1.00; specificity 0.92; positive predictive value 0.83; negative predictive value 0.99) are even better than those obtained in a similar study in the Mexican population. Conclusions: Considering that the MCHAT R/F is a parent-reported instrument, the Guayaquil Spanish version of the ADEC (ADEC-GU) seems to be a suitable instrument to be used in a complementary way as a second-level screening instrument for autism, before resorting to a full diagnostic process.