Increasingly, states and districts are combining student growth measures with rigorous, rubric-aligned teacher observations in constructing teacher evaluation measures. Although the student growth or value-added components of these measures have received much research and policy attention, the results of this study suggest that the data generated by high-quality observation systems have potential to inform principals' use of data for human capital decisions. Interview and survey data from six school districts that have recently implemented new evaluation systems with classroom observations provide evidence that principals tend to rely less on test scores in their human capital decisions. The consistency, transparency, and specificity of observation data may provide benefits for principals seeking to use these data to inform their decision making.