There are reasons why teaching behavior should be assessed, including (1) upgrading teacher education, (2) gaining insights into the learning of both teachers and children, and (3) studying social interactions. Two means of assessing teacher ability are quantification of teacher behavior by the use of rating scales, behavioral categories, etc., and participant observation (PO). The first, assessment by instrument, confounds the effects of too many interacting variables for the instrument to reliably represent the effects of teacher behavior. In the PO method, very well qualified and trained people are the assessing instrument. Observer judgment and observer influence upon the classroom situation are present, but if the observer is well qualified and well trained, as he must be for the success of the method, the data obtained should be more reliable and more relevant. Filming the classroom situation can also be used and adds much to the assessment process. The PO approach was tested on selected Head Start and elementary school classes. The data analysis from this testing is incomplete. It has been found, however, from a combined PO and filming of suburban and inner-city (Hartford, Connecticut) elementary classes, that suburban classes are uniformly superior to inner-city classes. (WD)