This thesis is written in the field of special needs education and developmental psychology at Institute of Education, University of Oslo. The study investigates measurement properties as well as suitability of measures that target language and social communication in children with Autism Spectrum Disorder (ASD). An overarching aim was to evaluate possible candidate measures for use in high-quality studies on populations of children with ASD. The three papers presented in this study use data from a randomized controlled trial conducted at Oslo University Hospital. The sample consists of 61 children diagnosed with childhood autism, and their respective parents and preschool teachers. The first two papers used baseline data to compare inter-rater reliability and the concurrent validity of language measures. The third paper compared two measures of social communication using baseline and post-intervention data to investigate how the choice of various measures may influence the outcomes of intervention studies. The results from the two first articles indicate that parents and preschool teachers have a similar view of children’s language levels and that report-based assessment of language may be an alternative if direct assessment is not available. Further, both report-based assessment and direct tests of language seems to be working very well to gather information about language in children with ASD. The aim of the third article was to evaluate in what way outcome measures that largely overlap with intervention targets “picks up” on effect compared to measures assessing broader aspects of functions. There seems to be an attenuation of effects when using a broader outcome measure compared to a measure that is proximal to the content of the intervention. This study adds to the body of knowledge that considers advantages and disadvantages of the use of measurement tools.