In reptiles, administration of opioid drugs has yielded unexpected results with respect to analgesia. The aims of this study were to assess the pharmacokinetic/pharmacodynamic (PK/PD) properties of tramadol and its active metabolite M1 and to evaluate the effect of the renal portal system on the PK/PD parameters in yellow-bellied slider turtles. Turtles (n = 19) were randomly assigned to four treatment groups, according to a masked, single-dose, four-treatment, unpaired, four-period crossover design. Group A (n = 5) received a single i.m. dose of tramadol (50 mg/mL) at 10 mg/kg in the proximal hindlimb. Group B (n = 5) received the same i.m. dose but in the forelimb. Groups C (n = 5) and D (n = 4) received a single i.m. injection of saline (NaCl 0.9%) of equivalent volume to the volumes of tramadol injected in the hind- and forelimb, respectively. Groups were rotated (1-month washout period) until the completion of the crossover study. Tramadol plasma concentrations were evaluated by a validated HPLC-FL method. An infrared thermal stimulus was applied to the plantar surface of the turtles' hindlimbs to evaluate the thermal withdrawal latency (TWL). The two PK profiles of tramadol differed in the first 2 h following administration, but overlapped in the elimination phases. The metabolite M1 was formed in both the treatment groups, showing similar pharmacokinetic trends, although the amount of M1 was significantly higher (20%) in the hindlimb vs. forelimb group. Turtles given tramadol in the hind- and forelimb showed a significant increase in TWL over the periods of 0.5-48 and 8-48 h, respectively. The calculated % maximal possible response (% MPR) was low (about 24%). The PK/PD correlations between M1 plasma concentrations vs. % MPR appeared to show a counterclockwise hysteresis loop shape.