This paper reports on secondary analyses of data collected in March 1978 by the Pennsylvania Educational Quality Assessment Program, which was designed to examine television viewing and the amount of school assigned homework in relation to student cognitive and noncognitive outcomes. Also examined were television viewing and homework patterns for groups categorized by sex, race, type of community, stability of home residence, parental educational level, size of family, and expectation level; data used were derived from 90,000 students in grades 5, 8, and 11 in 750 schools participating in a statewide assessment. Related studies from the states of Rhode Island and Texas are compared to this study. A student level analysis involving television indicates that there is little evidence of a meaningful relationship with cognitive or noncognitive achievement. In the school level analysis, where inferences refer to characteristics of schools rather than individuals, a strong negative relationship between television viewing and cognitive achievement was indicated. (Author/RAO)