Inventory inaccuracy remains a critical issue in the Peruvian textile industry, driven by manual tracking and lack of real-time visibility. This study aimed to improve finished-goods inventory accuracy in a jeans-manufacturing SME through the integration of Radio Frequency Identification (RFID), Value Stream Mapping (VSM), and discrete-event simulation. A time study over 14 days supported the development of current and future-state models using Arena Simulator. RFID gates and handheld readers were incorporated in the improved design. Results showed an increase in inventory record accuracy from 85% to 95%, an 81% reduction in registration time (from 2.6 h to 0.49 h), and a 35% reduction in lead time. The intervention also enabled 100% compliance with the daily production plan. This research fills a gap in the literature on RFID implementation in Latin American textile SMEs and offers a replicable model for industries seeking sustainable improvements in operational visibility, labor efficiency, and customer service performance.