Based on National Crime Panel (NCP) city surveys, a sample of from 16,000 to 18,000 burglaries was selected for the study. The sample was divided into categories on the basis of wealth, and the model was estimated separately for each category. Victimization data for 1973 and 1974 were used in the estimations. Findings indicated that loss and property damage were the major influences in the reporting decision, regardless of income level. The magnitude of the effect increased as either the level of the loss or the level of property damage increased, thus increasing the probability that the household would report the incident. Insurance had a positive effect on reporting, but age of household head and rent had a relatively weak influence. Overall, this estimated model was capable of accurately predicting reporting behavior in four cities not used in the primary estimation. Data tables, four references, and appendixes presenting a glossary of variables and NCP questions used to produce variables are furnished.