Description: Although several guidelines are available for designing observational procedures in both basic and applied settings, few researchers have experimentally examined and compared different direct observation methods. Such methods may have a direct impact on practitioners' ability to effectively assess strengths and challenges, set treatment goals, adjust intervention procedures, and monitor progress. The current study compared 1 and 5 min observations to 10 min observations throughout baseline and intervention phases of a parent training program for toddlers with autism. Results showed similarities with regards to variability, level, and trend in the 5 and 10 min data samples; however, clear differences were seen in the 1 min data sample, which typically showed very low occurrences of responding and displayed steady and flat trends. The findings have implications for the development of time-efficient direct observation procedures utilized in parent training programs for children with autism.