The details of quantitative film autoradiography for tritium using tritium plastic standards were examined with respect to 3 issues: tritium tissue equivalent (TE) calibration; correction of autoabsorption differences for gray and white matter; and the use of carbon-14 standard sources as a method for quantifying tritium tissue images. Both 3H-tissue and 3H-plastic sources produced linear log-log relationships of 3H-concentration (nCi/mg source weight) vs source optical density (OD) over a reproducible OD range (0.080-0.800). The curves for both 3H-tissue and 3H-plastic were parallel; uncalibrated 3H-plastic standards exhibited a 4-fold higher autoabsorption over 3H-tissue sources for OD values in the linear range. Using chloroform extraction of brains from rats treated with either [14C]deoxyglucose ([14C]DG) or [3H]deoxyglucose ([3H]DG), we found neither isotope loss nor redistribution after defatting (30% reduction of tissue dry weight). After chloroform extraction, the OD values from both gray and white matter structures containing carbon-14 were unaltered. Gray matter OD values increased by 28.7 +/- 5.6% (mean +/- S.D.) in structures containing tritium; white matter structures containing tritium exhibited a 115.9 +/- 29.3% increase in OD after chloroform extraction. The increase in OD after chloroform extraction was a fixed percent for any given tritium OD value from unextracted tissue when the value was within the linear range of 0.080-0.800 OD units. The magnitude of the higher white matter autoabsorption for tritium was confirmed using tritium impregnated cow brain pastes of variable gray/white mixtures. Chloroform extraction of tissue from [3H]DG treated rats was therefore a suitable procedure for direct correction of regionally heterogeneous autoabsorption of tritium. Finally, the rates of image generation for tritium and carbon-14 sources were compared. The rate of increase of OD with increasing exposure time was found to be equal for 3H-tissue and 3H-plastic images; sources of carbon-14 in plastic, however, exhibited more accelerated rates of image generation when compared to tritium sources (i.e. 3H- and 14C-images did not covary with exposure time). The effect of non-covariance on tritium TE calibrated standards was the overestimation of OD values for 14C-plastic standards with increasing times of exposure (comparison of 4 week images to 1 week images showed errors of 35-40%). Use of carbon-14 sources to quantify tritium-generated images therefore required recalibration of 14C-plastic for all exposure times of interest.