The fashion apparel industry is too diverse, volatile and uncertain due to the fast changing market scenario. Forecasting demands of consumers has become survival necessity for organizations dealing with this field. Many traditional approaches have been proposed for improving the computational time and accuracy of the forecasting system. However, most of the approaches have over-looked the uncertainty existing in the fashion apparel market due to certain unpredictable events such as new trends, new promotions and advertisements, sudden rise and fall in economic conditions and so on. In this chapter, an intelligent multi-agent based demand forecasting and replenishment system has been proposed that adopts features from nature-inspired computing for handling uncertainty of the fashion apparel industry. The proposed system is inspired from the group hunting behaviour of crocodiles such as they form temporary alliances with other crocodiles for their own benefit even after being territorial creatures.