Over a period of time, experiments were conducted to estimate the shelf life of a product. Each trial used a combination of a temperature level and an additive concentration that was used to inhibit spoilage. The policy was to terminate each trial after 270 days, even if the product sample had not yet failed. Particularly at the lower temperatures, some trials ended before the product sample reached the failed state. No records were kept on the number of unfailed samples. Thus the resulting data were truncated. This paper describes the analysis of the resulting data and the methods that were used to estimate the shelf life distribution of the product.