Dissipation of seven pyrethroid insecticides under field conditions was evaluated on green beans, zucchinis, and peppers grown in experimental greenhouses (Almería, Spain). Pyrethroid residues were determined by high performance liquid chromatography using continuous on-line post-elution photoirradiation with fluorescence detection after dichloromethane extraction and cleanup on florisil phase cartridges. Mathematically defined decline curves were established by determining optimal relationships between pyrethroid residues and time. Different models were used to find these curves. The 1st-order model achieved the best adjustment to the experimental data in 42.9% of cases. The RF (root function) 1st-order model was the best in 33.3% of times. Each of the 1.5th- and 2nd-order models provided the best adjustment in a 9.5% of the cases. Finally, the RF 1.5th-order model was the most appropriate in only 4.8% of cases. Half-life times for these three vegetables were estimated from the optimal models. The preharvest intervals for the residues in these three vegetables was obtained, taking into account the maximum residue levels established by the existing legislation. They were all lower than the ones specified by the makers of commercial formulates, which ensures a safe enough consumption.