How to Use CBD Oil

• Oral bioavailability: Absolute oral bioavailability on an empty stomach ~6% (an estimate; no detailed study measuring it directly has been published). No study has measured absolute bioavailability by the oral route (2018 systematic review). (Perucca & Bialer 2020, PMID 32504461, Millar 2018, PMID 30534073)
• First-pass effect: ~70–75% of the absorbed dose may be metabolized in the liver before it reaches the systemic circulation (an estimate). (Perucca & Bialer 2020, PMID 32504461)
• Taken with food (high-fat): With a high-fat, high-calorie meal, Cmax rises ~5-fold and AUC ~4-fold; variability decreases (Taylor 2018: Cmax 4.85-fold, AUC 4.2-fold). (FDA label 12.3, EMA SmPC 5.2, Taylor 2018, PMID 30374683)
• Taken with food (low-fat / milk): With a low-fat meal, Cmax ~4-fold and AUC ~3-fold; with cow’s milk, Cmax ~3-fold and AUC ~2.5-fold. EMA recommends keeping the timing relative to meals consistent (always fed or always fasting). (EMA SmPC 5.2)
• Taken with food (meta-analysis): Food increases how much CBD reaches the blood 2–5-fold. (Jang 2024, PMID 38850302)
• With alcohol: Alcohol increases CBD AUC by 63% (Cmax by 93% on the FDA label). (EMA SmPC 5.2, FDA label 12.3)
• Time to peak (Tmax): 2.5–5 hours at steady state; ~4–5 hours after a single dose. (FDA label 12.3, Taylor 2018, PMID 30374683)
• Half-life: Terminal 56–61 hours (twice daily, 7 days); effective half-life 10–17 hours. (FDA label, EMA SmPC, Taylor 2018, PMID 30374683)
• Reaching steady state: 2–4 days with twice-daily dosing (Taylor 2018: ~2 days). (EMA SmPC 5.2, Taylor 2018, PMID 30374683)
• Dose–exposure relationship: Less than dose-proportional increase (Cmax slope 0.73; AUC 0.64); absorption is saturable. (Taylor 2018, PMID 30374683, EMA SmPC)
• Protein binding: >94%. (FDA label 12.3)
• Metabolism: In the liver by CYP2C19 and CYP3A4; UGT1A7, UGT1A9, UGT2B7. Active metabolite 7-OH-CBD; main circulating metabolite 7-COOH-CBD. (FDA label, Taylor 2018, PMID 30374683)
• Excretion: Mainly in feces; a small amount through the kidneys. (FDA label)
• Formulation differences: After 100 mg of oral CBD, urine Cmax was 1274 ng/mL with Epidiolex, 776 with capsules and 151 with syrup. (Sholler 2022, PMID 34089060)
• Sublingual drops and capsules: No difference in blood levels between sublingual drops and capsules (the drops are most likely swallowed). (Johnson 2024, PMID 37733294)
• Low blood levels even at high doses: After 800 mg of oral CBD, the median blood level at 130 minutes was 5.0 µg/L; even at 800–1600 mg, blood levels were in the µg/L range. (Schneider 2022, PMID 34086631)
• Low-dose products: In studies of low doses (≤175 mg/day), plasma levels were 5–100 times lower than with a 20 mg/kg Epidiolex dose. (Warren 2026, PMID 42473010)
• Liver impairment: In moderate/severe hepatic impairment, exposure rises 2.5–5.2-fold. (FDA label 8.6, 12.3)

In the sources, the "start low, go slow" principle appears only under a doctor’s supervision and in research protocols; this page contains no personal dose recommendation. For the approved CBD medicine, the starting dose and the steps for increasing it are written in the medicine’s prescribing information and are set by a doctor. An expert consensus written for medical cannabis in chronic pain also suggests starting at a low dose and increasing it every few days; this is expert opinion only and is meant for use under a doctor’s supervision. In clinical trials, too, participants started at a low dose and were brought up to the study’s target dose over several weeks; these are research protocols, not individualized recommendations. The upper limits that agencies have calculated for food supplements (EFSA in the European Union, the FSA in the UK) are safety limits, not plans for increasing the dose. Your doctor decides, together with the medicines you take, whether CBD is right for you and at what dose.

FDA, prescribing information of the approved CBD medicine §2, Bhaskar 2021, PMID 34215346, Millar 2019, PMID 31222854, FSS/FSA 12.10.2023, www.efsa.europa.eu

These limits are for CBD sold as a food supplement; they are not the medicine doses used to treat diseases.

• EFSA (2026): 0.0275 mg/kg/day · 2 mg/day, Provisional safe level of 0.0275 mg/kg body weight/day ≈ 2 mg/day for a 70 kg adult. The value was derived by applying a total uncertainty factor of 400 (100 × 2 × 2) to a reference point calculated from liver findings in rats (BMDL10 11 mg/kg/day). Only for supplements of ≥98% purity that contain no nanoparticles, are produced safely and have had genotoxicity ruled out. Safety could not be established for people under 25, people who are pregnant or breastfeeding, and people taking medicines. Uncertainties remain about the liver and the nervous, endocrine, reproductive and immune systems. This is a food/supplement value; it should not be compared with prescription treatment doses. (www.efsa.europa.eu, EFSA Journal 2026, 24(2):e9862, doi:10.2903/j.efsa.2026.9862, full text PMC12884199 (accepted 16.12.2025, published 09.02.2026))
• EFSA NDA Panel (September 2026) — synthetic CBD: 0.026 mg/kg/day · 2 mg/day, Novel food opinion on chemically synthesized CBD of ≥98% purity: found safe under the proposed conditions of use — only as a supplement dissolved in MCT oil, at most 2 mg a day, for adults over 25 who are not pregnant or breastfeeding; safety could not be established in people taking medicines. The value was derived from a liver finding in female rats (BMDL10 10.5 mg/kg/day) with an uncertainty factor of 400. The opinion concerns a single application; the decision to authorize the product lies with the European Commission. (EFSA NDA Panel 2026, doi:10.2903/j.efsa.2026.10154, PMID 42761670)
• EFSA NDA Panel (2022): , The safety of CBD as a novel food cannot be established at this time; effects on the liver, the digestive tract, the endocrine and nervous systems and psychological effects need to be clarified; animal studies showed significant reproductive toxicity. Its effect on drug metabolism and genotoxicity are also among the data gaps. (EFSA Journal 2022, 20(6):7322, doi:10.2903/j.efsa.2022.7322, full text PMC9172591))
• UK FSA + FSS (2023): 10 mg/day, Healthy adults should limit CBD from food to 10 mg a day (≈ 4–5 drops of a 5% oil); the previous advice (February 2020) was 70 mg/day. Children, people who are pregnant or breastfeeding or planning a pregnancy, and people taking medicines without having consulted a doctor should not take CBD. Reason: adverse effects on the liver and thyroid with long-term use (FSS press page and FSA guidance page; reversible once stopped). (FSS/FSA press release 12.10.2023, www.foodstandards.gov.scot, www.gov.uk)
• UK ACNFP + COT (2023): 0.15 mg/kg/day · 10 mg/day, Provisional acceptable daily intake of 0.15 mg/kg/day ≈ 10 mg/day at 70 kg; ≥98% pure CBD, by mouth only. Main concerns: somnolence, liver; possible drug interactions above 70 mg/day (cannot be ruled out at lower doses). According to the committees’ 2025 summary report, the advice rests mostly on liver, thyroid, adrenal gland and kidney findings; the most sensitive effects are on the liver and thyroid. (ACNFP/COT joint position paper 12.10.2023, www.gov.uk, ACNFP–COT CBD subgroup summary report 17.10.2025, www.gov.uk)
• US FDA (2019): , 'Strongly' advises against any use of CBD, THC and marijuana during pregnancy and breastfeeding. (FDA consumer page 16.10.2019, www.fda.gov)
• ASCO guideline (2024) — cancer: 300 mg/day, In adults with cancer, oral CBD at ≥300 mg a day is not recommended for symptom burden outside a clinical trial (no proven efficacy; risk of reversible liver enzyme abnormalities; evidence quality low, recommendation weak). This is a professional guideline recommendation, not an official food limit. (PMID 38478773, doi:10.1200/JCO.23.02596)

• 84 products bought online (31 companies): 30.95% accurately labeled (±10%), 42.85% contained more CBD than labeled, 26.19% less. THC was found in 18/84 products (21.43%), up to 6.43 mg/mL. (Bonn-Miller 2017, PMID 29114823)
• After 100 mg of pure CBD (oral/vaporized), only 1 of 218 urine samples was positive at the 20 ng/mL screening cutoff; none exceeded the 15 ng/mL confirmation cutoff. (Spindle 2020, PMID 31682266)
• CBD-dominant product containing THC (100 mg CBD + 3.7 mg THC, vaporized): four samples positive on confirmation testing (THC-COOH >15 ng/mL); in a second study, 6 samples from 3 of 18 people were ≥15 ng/mL. (Spindle 2020, PMID 31682266, Sholler 2022, PMID 34089060)
• MCT oil with 100 mg CBD twice a day for 14 days (60 healthy adults, groups of 10): in every group receiving THC (0.5–3.7 mg THC), at least one person had a positive urine test; in the CBD-only group, one urine sample was positive. THC in the blood was very low; no difference was seen in intoxication or attention tests. One of the authors is an employee of a CBD company. (Wolinsky 2026, PMID 40704705)
• There is no evidence that pure CBD converts to Δ8/Δ9-THC by any route or in any formulation. (Sholler 2022, PMID 34089060)
• No positive urine tests with topical products; however, THC ≥2 ng/mL in oral fluid in 9 participants. (Zamarripa 2024, PMID 38217086)
• Pure oral CBD (15–1500 mg) did not produce THC-positive results on two roadside saliva tests. (McCartney 2022, PMID 34412166)
• In users of a broad-spectrum CBD product (150 mg/day, 10 weeks), CBG and CBDV, which are prohibited under WADA’s 2026 list, were detected in urine (42% and 68% before exercise, 74% and 84% after); THC was not detected. CBDV was not listed on the product’s certificate of analysis. (Gillham 2026, PMID 40920736)
• 'Full-spectrum' (other cannabinoids, including THC), 'broad-spectrum' (THC is claimed to have been removed) and 'isolate' (pure CBD) are common industry terms; no binding regulatory definition was reported in the sources reviewed. Regulatory safety values were derived only for ≥98% pure CBD. ACNFP/COT groups products as A (≥98% pure CBD), B (CBD + cannabinoid mixture) and C (natural hemp extract) (secondary source). (EFSA 2026, ACNFP/COT 2023)
• In Australia, 17 of 58 (29%) unregulated cannabis products used in children with epilepsy had ethanol or isopropanol above the limit (5000 ppm); one sample had bifenthrin (a pesticide) at 4.9 times the limit; no sample exceeded heavy metal limits (51 samples). (Suraev 2022, PMID 34954507)
• The content and purity of over-the-counter CBD products used in pain studies are a problem; the content on the label may not match the actual content. (Moore 2024, PMID 37863344)
• Dependence: Epidiolex is not a controlled substance; no withdrawal symptoms were seen after 28 days of 1500 mg/day. (FDA Epidiolex label 9.x)

In Turkey, the sale of cannabis-derived products is governed by a regulation dated January 31, 2026. Details: current regulation.