Researchers found strong correspondence between yearly deviations in UCR crime rates and NCS victimization rates for both burglary and robbery. Differences between the two data series primarily involved contrasting trends, although there was some evidence that UCR and NCS crime rate trends have converged in recent years. Ex post forecasts revealed that UCR/NCS relationships estimated from 1973-1985 data continued through 1986 and 1987. While UCR rates in 1986 were somewhat influenced by unusual increases in the proportion of crimes reported to the police that year, changes in crime reporting for the period as a whole had little effect on UCR burglary and robbery rates. The researchers conclude that there is strong consistency between UCR and NCS data sources on variations in crime rates over time. This consistency does not mean that raw crime rate measures provided by each data series are similar, but rather that the two series are systematically related to each other over time such that the value of one series can be estimated with reasonable accuracy from the value of the other. Factors affecting crime reporting rates are discussed in relation to reconciling data from different sources. 36 references, 4 tables, and 1 figure