This document discusses predicting product lifetime with small sample sizes. It provides a method using cumulative probability distributions and compares data from 10 sample failures tested to failure using a Weibull distribution and lognormal distribution. The Weibull distribution provides a more conservative prediction of lifetime at the 1% failure rate compared to the lognormal distribution, predicting failure would occur 11,000 hours earlier. Using the wrong distribution can result in an overly optimistic lifetime prediction and unexpected early field failures. The Weibull distribution is recommended for small sample size prediction to avoid this.