Sibling data drawn from the Kalamazoo Brothers sample are used in this report to assess the adequacy of conventional sociological variables for measuring family background, to estimate the overall effects of family background on test scores, education, occupational status, and earning, and to control family background when estimating the effects of test scores and education. Traditional socioeconomic variables are found to be imperfect measures of background. The correlations between brothers' test scores, educational attainment, occupational status, and earnings are substantially higher than would be predicted on the basis of measured socioeconomic background alone. Nevertheless, the differences between brothers on measures of economic success are large relative to differences among men in general. Differences in family background explain less than one quarter of the variance in earnings. Biases in the effects of education on occupational status and earnings due to background are not fully eliminated by controlling only measured background variables. Controlling both sibling test score differences and common family background suggests a 30% bias in the education-occupation coefficient and a 54% bias in the education income earnings coefficient. Controlling common family background does not, however, substantially reduce the effect of test scores on income earnings. (Author/AM)