The long-term behavior of the herbicide atrazine and its metabolites in the environment is of continued interest in terms of risk assessment and soil quality monitoring. Aqueous desorption, detection, and quantification of atrazine and its metabolites from an agriculturally used soil were performed 22 years after the last atrazine application. A lysimeter soil containing long-term aged atrazine for >20 years was subdivided into 10 and 5 cm layers (at the lysimeter bottom: soil 0-50 and 50-55 cm; fine gravel 55-60 cm depth, implemented for drainage purposes) to identify the qualitative and quantitative differences of aged (14)C-labeled atrazine residues depending on the soil profile and chemico-physical conditions of the individual soil layers. Deionized water was used for nonexhaustive cold water shaking extraction of the soil. With increasing soil depth, the amount of previously applied (14)C activity decreased significantly from 8.8% to 0.7% at 55-60 cm depth whereas the percentage of desorbed (14)C residues in each soil layer increased from 2% to 6% of the total (14)C activity in the sample. The only metabolite detectable by means of LC-MS/MS was 2-hydroxyatrazine while most of the residual (14)C activity was bound to the soil and was not desorbed. The amount of desorbed 2-hydroxyatrazine decreased with increasing soil depth from 21% to 10% of the total desorbed (14)C residue fraction. The amount of (14)C residues in the soil layers correlated well with the carbon content in the soil and in the aqueous soil extracts ( p value = 0.99 and 0.97, respectively), which may provide evidence of the binding behavior of the aged atrazine residues on soil carbon. The lowest coarse layer (55-60 cm) showed increased residual (14)C activity leading to the assumption that most (14)C residues were leached from the soil column over time.