Section 3.1 mentioned that (a) foods are heterogeneous, (b) the chemical reactions involved in the deterioration of foods have complex kinetics, (c) rheological and textural properties are not simply modeled, (d) biological changes continue during food storage, and (e) the deterioration of a food during storage is often the result of a number of simultaneous processes. All this leads to difficult practical and statistical issues involved in accelerating the lives of complex products like most foods. The usual procedure is for information from tests at high levels of one or more accelerating variables (e.g., temperature, moisture, or light) to be extrapolated to obtain estimates of shelf life at lower, normal levels of the accelerating variable. Such testing saves much time and money. Pitfalls of accelerated tests will be discussed in Section 7.5.