This paper models the supply of perishable goods within a randon framework. Perishability affects a large group of goods usually traded in the economy such as fruits and vegetables, newspapers, medicine drugs, a.s.o.. Surprisingly, one cannot find in the literature a decision model for suppliers that takes into account the specificity of this kind of goods. The suppliers guess their demand by choosing a probability density function, one at each price level. Then they choose optimal supply functions maximizing their expected profits. Examples of the optimal solution are given for some known demand distribution functions like Pareto and Weibull. The autarchic model is then extended to include nonprice competition among the sellers. Each seller chooses the supply curve that maximizes his expected profit, conditioned by the event that competitors’ markets are in equilibrium. The supply of rivals affect the sales for certain to loyal clients, but not the random sales. The autarchic model is then used to analyze the green-pepper market in Rio de Janeiro(1994/7-2000/11). The results give consistency to the rational hypothesis of the model
