Business tendency surveys are a commonly accepted instrument for the assessment ofthe current business cycle course. Most of these surveys rely on qualitative questionsabout the current situation of the firms and about heir expectations for the next months. This paper analyzes whether qualitative questions about employment expectations are useful to assess actual employment changes. In Germany the Ifo Institute is specialized on business surveys. The German Ifo data are investigated using three different approaches: Smoothing techniques help to date turning points in the course of the series. Error correction models are used to analyze the general lead/lag relations and Probit models are used to estimate a threshold for the survey based indicator which helps to differentiate between an increase and an decrease in employment. All three methods indicate that the employment expectations form a leading indicator of actual employment changes.