Adaptation of horizontal vestibulo-ocular reflex (HVOR) provides an experimental model for cerebellum-dependent motor learning. The HVOR gain, measured by 10 deg (peak-to peak amplitude) and 0.16-0.5 Hz turntable oscillation in dark, was around unity. The gain-up and gain-down HVOR adaptation was induced by 1 h of sustained out-phase and in-phase 10 deg-0.33 Hz combined turntable-screen oscillation in the light, respectively. To examine the role of long-term depression (LTD) of parallel fiber-Purkinje cell synapses, we intraperitonially applied T-588 or nimesulide, which block the induction of LTD in vitro or in vivo preparations, 1 h before the test of HVOR adaptation. T-588 (3 or 5 mg/kg body weight) did not affect nonadapted HVOR gains, and impaired both gain-up and gain-down HVOR adaptation. Two previous studies (Welsh et al.,2005; Schonewill et al., 2011) reported that T-588 did not alter rodent motor learning by oral or intrapertoneal applications, but the present results suggest that it is due to the difference of method of application or dose of drugs used, and species. Nimesulide (3 or 6 mg/kg) did not affect nonadapted HVOR gains, and impaired gain-up HVOR adaptation dose-dependently; however, it very little affected gain-down HVOR adaptation. We discuss on the neural mechanisms underlying the different actions of two drugs on motor learning.