Over the last decade, the usage of short message services (SMS) as one of the vital communication services on mobile devices has grown. The growth of using this service has correspondingly increased the number of attacks on mobile devices such as SMS Spam. SMS spam is a concern to telecommunications service providers as they annoy the subscribers and cause them to loose commercial. Most current researches have attempted to detect SMS spam using different classifiers. In this paper, we propose a new method that focuses on binary particle swarm optimizationbased fuzzy rules selection for detecting SMS spam messages. First, we extract significant features from the SMS spam dataset. Then, a set of fuzzy rules based on the extracted features is generated. Finally, a binary particle swarm is suggested for picking the more powerful fuzzy rules that reduce the complexity and improve the model's performance. The SMS Spam benchmark dataset was used in the experiment. The attained results of the proposed model provide a recall of 98.8%, precision of 90.8%, F-measure of 94.6% and accuracy of 98.5% that indicate the proposed model can be a promising for detecting SMS spam.