The article explores how non-stationary cycles affect unit root tests commonly used in statistical analysis. The study focuses on various statistical tests and their accuracy when dealing with cycles. The results show that some tests are reliable, like the augmented Dickey-Fuller test, while others are not as trustworthy. Additionally, the study confirms that certain methods used to determine the order of augmentation in the Dickey-Fuller test are still valid even in the presence of non-stationary cycles.