One of the most used approaches for providing recommendations in various online environments such as e-commerce is collaborative filtering. Although, this is a simple method for recommending items or services, accuracy and quality problems still exist. Thus, we propose a dynamic multi-level collaborative filtering method that improves the quality of the recommendations. The proposed method is based on positive and negative adjustments and can be used in different domains that utilize collaborative filtering to increase the quality of the user experience. Furthermore, the effectiveness of the proposed method is shown by providing an extensive experimental evaluation based on three real datasets and by comparisons to alternative methods. We propose a method that improves collaborative filtering recommendations.It may use either a static or a dynamic multi-level approach.It is based on positive and negative adjustments of the users similarity values.Both approaches have been experimentally evaluated using three real datasets.Our approaches produce results of better quality when compared to alternatives.