We use a mixture regression model to identify segmentation in the Israeli labor market, and propose a new method for assigning workers to simulated segments. We identified a lowwage segment and a high-wage segment, as well as a third segment with a large wage variability that we interpret as “noisy” observations. We found quantitatively small but qualitatively reasonable differences in workers’ characteristics between the low-wage and high-wage segments, while the coefficient differences were much larger, indicating that much of the wage disparity in Israel is due to unobserved factors rather than to observable characteristics. Some policy-relevant insights are derived.