An empirical study was conducted to explore text editing performance in an actual work environment. The text editing performance of 12 experts and 24 novices was studied across several benchmark tasks using either a command-driven, PC-based or hard-wired text editing system. Experts were tested for performance and functionality; novices were tested for learning. Additionally, the keystroke-level model was applied to the performance tasks and the results compared to actual observations. The results indicated that the methodology of T. Roberts (unpub. Ph.D. thesis, Stanford Univ., CA, USA, 1980) could be easily applied to a work environment and provided useful information for the evaluation of text editing systems. The study identified several areas of the keystroke-level model that could be modified to provide a more accurate assessment of text editing performance.< <ETX xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">></ETX>