Heart Rate Variability (HRV) has been garnering a lot of attention from medical researchers and biomedical engineers due to its ability to expose crucial information about the status of the nervous system and the health of the human heart. Although time domain analysis of a HRV signals can yield a wealth of information, frequency domain analysis has been gaining in popularity. This is mainly due to the identification of distinct frequency bands that reflect specific components of the nervous system. Nonetheless, signal artifact can severely distort the extracted time and frequency domain parameters alike and thus rendering the information obtained from the signal unusable. In this paper, we propose the use of a Windowed Impulse Rejection (WIR) based artifact detection algorithm. Our performance evaluation demonstrated that our method performs with a higher level of accuracy than its competitors. Also, in terms of complexity, it outperformed the Moving Average algorithm.