Heart sounds contain important information related to heart health. Normal heart sounds produce sound patterns that are different from abnormal ones. Various digital signal processing methods have been used to differentiate these heart sound signals and the most usual methods are wavelet analysis, entropy analysis, or a combination of both. In this study, the fractal dimension was used to classify normal heart sounds and murmur sounds by using the box-counting fractal dimension (BCFD), Katz fractal dimension (KFD), Sevcik fractal dimension (SFD), and Higuchi fractal dimension (HFD) as the heart sound features. The highest accuracy reached 100% using SFD as a feature and KNN as a classifier. These results were tested on 50 normal heart sounds and 50 heart sounds murmurs with 3-fold cross-validation. The results showed that choosing the right fractal dimension can distinguish normal heart sounds and murmurs.