Methods for forecasting harvest yields have been improved considerably in the last 20 years with the development of new data survey (remote sensing) and statistical techniques. One of these methods, based on pollen release in the atmosphere, is especially important for anemophilous species such as olive. The aim of the present work is to use a different approach to forecast the olive harvest by considering the pollen variable as “endogenous” because it is involved in the consequential processes from the formation of pollen to fruiting, the complex of which determines, more or less, the final production. Unlike models built upon a single equation (multiple linear regression analysis), the proposed estimate, based on an incomplete system of equations, recovers the consistency associated with the inference of parameters while avoiding the errors of “over-estimation.” The study, based on 17 years of data considers the quantity of olive pollen monitored and the relative annual olive production in addition to climatic, agronomic, and pathological variables associated with production. The harvest forecast provides the possibility for planning and optimizing the various stages of olive production from cultivation to distribution, including sound management of the olive supply.