Tramadol hydrochloride is a centrally acting analgesic used for the treatment of moderate-to-severe pain. It has three main metabolites: O-desmethyltramadol (M1), N-desmethyltramadol (M2), and N,O-didesmethyltramadol (M5). Because of the frequent use of tramadol by patients and drug abusers, the ability to determine the parent drug and its metabolites in plasma and cerebrospinal fluid is of great importance. In the present study, a pharmacokinetic approach was applied using two groups of five male Wistar rats administered a 20 mg/kg dose of tramadol via intravenous (i.v.) or intraperitoneal (i.p.) routes. Plasma and CSF samples were collected at 5–360 min following tramadol administration. Our results demonstrate that the plasma values of Cmax (C0 in i.v. group) and area under the curve (AUC)0–t for tramadol were 23,314.40 ± 6944.85 vs. 3187.39 ± 760.25 ng/mL (Cmax) and 871.15 ± 165.98 vs. 414.04 ± 149.25 μg·min/mL in the i.v. and i.p. groups, respectively (p < 0.05). However, there were no significant differences between i.v. and i.p. plasma values for tramadol metabolites (p > 0.05). Tramadol rapidly penetrated the blood–brain barrier (BBB) and blood–cerebrospinal fluid barrier (BCSFB) (5.00 ± 0.00 vs. 10.00 ± 5.77 min in i.v. and i.p. groups, respectively). Tramadol and its metabolites (M1 and M2) were present to a lesser extent in the cerebrospinal fluid (CSF) than in the plasma. M5 hardly penetrated the CSF, owing to its high polarity. There was no significant difference between the AUC0–t of tramadol in plasma (414.04 ± 149.25 μg·min/mL) and CSF (221.81 ± 83.02 μg·min/mL) in the i.p. group. In addition, the amounts of metabolites (M1 and M2) in the CSF showed no significant differences following both routes of administration. There were also no significant differences among the Kp,uu,CSF(0–360) (0.51 ± 0.12 vs. 0.63 ± 0.04) and Kp,uu,CSF(0–∞) (0.61 ± 0.10 vs. 0.62 ± 0.02) for i.v. and i.p. pathways, respectively (p > 0.05). Drug targeting efficiency (DTE) values of tramadol after i.p. injection were more than unity for all scheduled time points. Considering the main analgesic effect of M1, it is hypothesized that both routes of administration may produce the same amount of analgesia.