Occupational monitoring data are typically in the form of air samples or biological samples. Air samples are more frequently available and often have been used to characterize personnel exposure in epidemiological studies. Air samples that are not specific to individual employees are easier and cheaper to procure than biological samples such as urinalyses. However, the correlation between concurrent air samples and urinalyses has not always been found to be strong. The purpose of this paper is to compare internal radiation doses for uranium workers estimated from air sample results with those estimated from urine sample results. The comparison was made on results associated with individuals who worked in a uranium processing and fabrication facility between 1950 and 1956. Independent lung dose estimates were made for individuals for whom both types of data were available for periods of 300-365 d. Plots of the data and statistical analyses failed to show evidence of correlation of any practical significance between the data generated by the two methods. A number of unquantifiable variables were enumerated for both types of estimates. It is concluded that within this study (1) only minimal correlation was demonstrable between lung doses based on air monitoring and doses based on urinalysis data because of the number of such variables affecting the measurements; and (2) general air data would not be the most useful measure of exposure upon which to base lung dose estimates.