L-DOPA-induced dyskinesia (LID) is a major complication of the pharmacotherapy of Parkinson's Disease. A model of LID has recently been described in rats with unilateral 6-hydroxydopamine (6-OHDA) lesions. In the present study, the model was used in order to compare the efficacies of some clinically available compounds that have shown antidyskinetic effects in nonhuman primate models of LID and/or in patients, namely, amantadine (20 and 40 mg/kg), buspirone (1, 2 and 4 mg/kg), clonidine (0.01, 0.1 and 1 mg/kg), clozapine (4 and 8 mg/kg), fluoxetine (2.5 and 5 mg/kg), propranolol (5, 10 and 20mg/kg), riluzole (2 and 4 mg/kg), and yohimbine (2 and 10 mg/kg). Rats were treated for 3 weeks with L-DOPA for an induction and monitoring of abnormal involuntary movements (AIMs) prior to the drug screening experiments. The antidyskinetic drugs or their vehicles were administered together with L-DOPA, and their effects were evaluated according to a randomized cross-over design both on the AIM rating scale and on the rotarod test. Most of the compounds under investigation attenuated the L-DOPA-induced axial, limb and orolingual AIM scores. However, the highest doses of many of these substances (but for amantadine and riluzole) had also detrimental motor effects, producing a reduction in rotarod performance and locomotor scores. Since the present results correspond well to existing clinical and experimental data, this study indicates that axial, limb and orolingual AIMs possess predictive validity for the preclinical screening of novel antidyskinetic treatments. Combining tests of general motor performance with AIMs ratings in the same experiment allows for selecting drugs that specifically reduce dyskinesia without diminishing the anti-akinetic effect of L-DOPA.