Computational thinking (CT) is a category of thought processes that allows framing and solving problems with computers, robots and other computational devices. The development of CT abilities through computational learning is increasingly vital to success in our technologically-oriented society. A major barrier to computational learning in schools is the current lack of validated instruments for assessing CT abilities in young children. We have developed a new CT assessment instrument called "TACTIC-KIBO" ( Tufts Assessment of Computational Thinking In Children - KIBO version). This instrument uses the KIBO programmable robot to evaluate the CT abilities of kindergarten and early elementary school children in classroom or research settings. TACTIC-KIBO rates overall CT ability into four levels by evaluating performance in seven CT categories that are based upon Bers' Developmentally Appropriate Seven Powerful Ideas of Computational Thinking (Bers 2018). In a pilot study, fifteen 5-7-year olds with various levels of past exposure to the KIBO robot were assessed with this new instrument. The children were also videotaped as they engaged in structured interactive play sessions with the KIBO robot to allow their CT skills to be rated independently by researchers with expertise in assessing CT. Results show a high correlation (r=0.895, p <0.001) between the ratings made with TACTIC-KIBO and expert assessments. TACTIC-KIBO could be easily administered and scored for children with a wide range of CT abilities in an average of 16 minutes per child. Expert ratings of interactive play sessions took longer to perform and were subject to inter-rater variability. Detailed analysis of participants' responses reveals potential areas in which TACTIC-KIBO can be improved in the future. Results of this pilot study are encouraging and provide support for the further development of TACTIC-KIBO as a practical assessment tool for educators and researchers to use when measuring CT abilities in young children.