Pollution of soil and groundwater by atrazine has become an increasing environmental concern in the last decade. A phytoremediation test using plastic pots was conducted in order to assess the ability of several crops and grasses to remove atrazine from a soil of low permeability spiked with this herbicide. Four plant species were assessed for their ability to degrade or accumulate atrazine from soils: two grasses, i.e., ryegrass (Lolium perenne) and tall fescue (Festuca arundinacea), and two crops, i.e., barley (Hordeum vulgare) and maize (Zea mays). Three different doses of atrazine were used for the contamination of the pots: 2, 5 and 10 mg kg−1. 16 days after spiking, the initial amount of atrazine was reduced by 88.6–99.6% in planted pots, while a decrease of only 63.1–78.2% was found for the unplanted pots, thus showing the contribution of plants to soil decontamination. All the plant species were capable of accumulating atrazine and its N-dealkylated metabolites, i.e., deethylatrazine and deisopropylatrazine, in their tissues. Some toxic responses, such as biomass decreases and/or chlorosis, were observed in plants to a greater or lesser extent for initial soil doses of atrazine above 2 mg kg−1. Maize was the plant species with the highest ability to accumulate atrazine derivatives, reaching up to 38.4% of the initial atrazine added to the soil. Rhizosphere degradation/mineralization by microorganisms or plant enzymes, together with degradation inside the plants, have been proposed as the mechanisms that contributed to a higher extent than plant accumulation to explain the removal of atrazine from soils.