This paper explores the extent to which goods follow systematic pricing patterns over their life cycle. The theoretical literature, and anecdotal evidence, often suggests that new products are often introduced at high prices which decline as the good ages while, older goods exit the market at a discount. We outline and apply a smoothing-spline approach to the estimation of life cycle price effects using data on two different types of goods (desktop and laptop computers, and personal digital assistants). We interpret these results within a simple conceptual framework and find evidence for the existence of significant life cycle pricing effects. This implies that hedonic pricing functions that exclude age are misspecified. Furthermore, in order to eliminate biases in price indexes, samples must be constructed carefully. Using a simulation we show that the bias introduced by the traditional match-model method may be non-trivial.