We assess biases due to measurement errors in structural equation models of the intergenerational transmission of socioeconomic status. Using data for nonblack males from the 1973 Occupational Changes in a Generation-II survey, we find that retrospective reports of status variables are as reliable as contemporaneous reports and that response errors are strictly random. When measurement errors are ignored, occupational returns to schooling are underestimated by about ten percent, the effects of some background variables are underestimated by as much as 19 percent, and residual variation in socioeconomic achievements is underestimated by as much as 16 percent.