Statistical time-to-failure analysis is a very powerful and versatile tool available to reliability engineers and statisticians for understanding and characterizing the failure risk and reliability of a component, device or system. Commonly applied methods of modeling time-to-failure involve fitting a parametric distribution, such as the Weibull probability function, using serial data on production or sales and incident data on units experiencing field failure since the launch of a product. When both the date of manufacture or sale and the date of incident are available from existing records, or can be easily ascertained by examination, the age of a failed unit can be determined exactly for purposes of analysis. Age censoring occurs, however, when one or both dates are missing-e.g., due to incomplete incident reporting or failure-induced physical damage to the unit. Excluding cases with incomplete date records involves a potentially significant loss of information in the time-to-failure analysis. We present several case studies to demonstrate how, in practice, such incidents can be treated as interval-censored observations in the time-to-failure analysis. Further, we evaluate the sensitivity of inferences to the inclusion of partially documented incidents to assess the value of this approach in practical applications.