Research on the relation between teacher behavior and pupil change has been an active field of inquiry for a number of years but the degree of agreement of results across various studies is not high. This paper is an examination of four studies seeking parallel bits of results which may be of use to classroom teachers. Research methodologies were parellel in that (a) pupil cognitive and noncognitive measures of classroom mean regressed gain from fall to spring were the dependent measures; (b) factor scores from two to four observation instruments were used as independent measures; and (c) nonlinear relations were tested in all studies, interactions in two. The greatest pupil gain was associated with (1) intermediate amounts of teacher control, the particular amount varying with task complexity; and (2) smaller amounts of work at high cognitive levels. Pupil socioeconomic status, more often than variables, interacted with classroom process. (RC)