Precision teaching emerged from O.R. Lindsley's pristine application of Skinner's natural science of behavior, with a focus on response rate measurement and free operant procedures. When applied with human learners in instructional settings, these first principles led to a series of developments framed in this paper as four kinds of ceilings that constrain or limit performance. Measurement-defined ceilings are the limits imposed on measurement sensitivity by use of percent correct and absence/presence methods of evaluation that lack standard units of measurement and ignore the time dimension. Removing measurement ceilings by using time-based measures leads to discovery of limitations imposed by procedures or materials that limit the pace at which behavior can occur, referred to in this article as procedure-imposed (or teacher-imposed) ceilings. By changing materials and procedures to allow more self-paced, unconstrained behavior, teachers encountered limits on the pace at which learners could perform, below those count per minute levels shown by people judged to be competent in the particular behavior being measured. These deficit imposed ceilings were remediated by identifying and building the frequencies of behavior components that must be combined and applied in more complex repertoires. Finally, the history of precision teaching has been, to a large degree, a search for behavior components which, if accelerated, could enable learners to break through what have been called handicap-imposed ceilings to achieve competent levels of count per minute performance. This article suggests a shift in the discussion among applied behavior analysts about precision teaching and fluency-based instruction from a focus on specific instructional procedures and recipes to the issue of measurement sensitivity and adherence to the underlying characteristics of Skinner's natural science, including response-rate measurement and free operant procedures.