Both policy makers and scholars frequently use monthly industrial production index series in their econometric predictions. In this paper, the stationarity properties of level, logarithm, seasonally adjusted level and seasonally adjusted logarithm of the Turkish industrial production index (TUIK) are investigated using both traditional unit root tests and unit root tests that have more power in small samples. Additionally, unit root tests that determine structural breaks endogenously are also employed. The traditional and relatively more recent tests seem to contradict each other Furthermore, there are structural breaks in the series. Therefore, model choices based on traditional unit root tests and policy implications derived from such models may not be correct. This study can help studies that employ the Turkish industrial production index in assessing the stationarity properties of the index.