CBD dosage guide: how much to take, how to measure it, and how to adjust

CBD serving sizes confuse people for one reason: the label gives a bottle total, the research gives milligrams per kilogram, and neither tells you where to start. This guide closes that gap. It shows the math behind "1500 mg per 30 mL," what the published human studies actually dosed, how food and format change absorption, and how to adjust over two weeks with a simple log.

How much CBD should a first-time user take?

The reason to begin small is not that CBD is fragile at higher amounts. A 2017 safety review notes that chronic use at doses up to 1500 mg per day has repeatedly been reported as well tolerated in humans, and that "the most commonly reported side effects were tiredness, diarrhea, and changes of appetite/weight" (Iffland and Grotenhermen, 2017). The reason is that you cannot judge what a serving does for you until you have held it constant for several days and written down what you noticed.

Published guidance on cannabis preparations in general says to begin at modest levels and that titration "should be undertaken slowly over a period of as much as two weeks" (MacCallum and Russo, 2018). Two weeks is long enough to see whether a steady serving changes anything about your routine, and short enough that you will not lose patience.

Body weight matters, but less than people assume at consumer servings. The prescription CBD product Epidiolex is dosed by weight at 2.5 mg/kg twice daily to start, which is 170 mg twice daily for a 68 kg (150 lb) adult (FDA prescribing information). Consumer products sit far below that range, so a 120 lb person and a 220 lb person will usually both begin at the low end and separate only as they titrate upward.

How do you calculate mg per dropper?

The dropper math is the same for every tincture. Write the concentration on the bottle in marker the day you open it: "50 mg/mL" is faster to use than "1500 mg / 30 mL." If the label gives a per-serving number instead, check that it matches the arithmetic. A 1000 mg bottle at 30 mL is 33.3 mg per mL; a label that calls one full dropper "50 mg" on that bottle is wrong.

Most droppers are marked at 0.25 mL, 0.5 mL, 0.75 mL and 1.0 mL. Use the marks, not the drop count, because drop size varies with the dropper tip, the carrier oil and the angle you hold it. If you must count drops, count how many drops your dropper produces in 1 mL once, write that number on the bottle too, and divide. TODO: verify the drops-per-mL of any specific dropper before publishing a per-drop figure.

Label accuracy is the hidden variable. In the JAMA sample, 42.85% of products carried less CBD than labeled and 26.19% carried more; tinctures were the least accurate format at 25.00% correctly labeled, and THC was detected in 21.43% of samples (Bonn-Miller et al., 2017). The federal research agency NCCIH makes the same point in plain words: "over-the-counter CBD products may contain more or less CBD than stated on their labels" (NCCIH). A current, batch-matched certificate of analysis is the only way to confirm the number you are dividing.

What does the research use as a dose range?

The prescription range is the best-documented. The label specifies a starting dosage of 2.5 mg/kg twice daily (5 mg/kg/day), an increase after one week to 5 mg/kg twice daily (10 mg/kg/day), and a maximum recommended maintenance dosage of 10 mg/kg twice daily (20 mg/kg/day) (FDA prescribing information). For a 70 kg adult that is 350 mg to 1,400 mg per day, taken under a prescriber's supervision with blood monitoring. Those numbers tell you what a pharmaceutical dose looks like; they are not a template for a tincture.

Single-dose laboratory work in healthy adults has used 300 mg. In a crossover study, 26 volunteers took 300 mg of CBD or placebo 30 minutes before an 8-hour sleep recording, and no sleep parameter differed between the two nights (Linares et al., 2018). That finding matters for dosing because it shows a large single serving did not produce a measurable sleep effect in people who already slept normally.

At the low end, the largest consumer trial to date randomized 1,793 adults to 4 weeks of capsules containing 15 mg of CBD or 5 mg of melatonin, alone or with 15 mg of CBN and 5 mg of CBC. Between 56% and 75% of participants across all arms reported a clinically important improvement in sleep quality, and the authors concluded the CBD effects "do not exceed that of 5 mg melatonin" (Saleska et al., 2023). The study was run by a company that sells trial services to supplement brands, which is a reason to read it as a tolerability and comparison study rather than proof of anything.

How long does a tincture take to absorb versus a gummy or capsule?

The common claim that a tincture "works in 15 minutes" is not supported by the pharmacokinetic literature. Holding oil under the tongue may let some CBD cross the oral mucosa, but the review's oromucosal data still show peaks measured in hours, not minutes (Millar et al., 2018). Most of a swallowed tincture, gummy or softgel follows the same path: stomach, small intestine, liver, then blood.

The prescription oral solution gives the cleanest numbers because it was studied under controlled conditions. Its label reports a time to maximum plasma concentration of 2.5 to 5 hours at steady state and a plasma half-life of 56 to 61 hours after twice-daily dosing for 7 days in healthy subjects (FDA prescribing information). That long half-life is why daily use builds up over the first week and why a two-week titration window makes sense.

Gummies and capsules add a dissolution step before absorption, so expect them to be at the later end of the oral range. The practical rule: take a tincture, gummy or capsule 1 to 2 hours before the moment you care about, keep the timing consistent, and judge the serving after several days, not after one evening.

Should you take CBD with food?

CBD is fat-soluble, which is why it is sold in carrier oils and why fat in a meal pulls more of it into circulation. The systematic review found the same direction in every study that tested it: "Plasma levels of CBD were increased when CBD was administered with food" (Millar et al., 2018). In one oromucosal spray study of 12 men, total exposure was 23.1 fed versus 4.5 fasted (h×ng/mL) and peak concentration 3.7 versus 1.2 ng/mL, roughly five and three times higher with food.

The label-level food effect was measured with 750 mg and 1500 mg doses of the oral solution, and the authors of the review conclude that "it may be advisable to administer CBD orally in a fed state" (Millar et al., 2018). Whether the same multiplier applies at a 25 mg gummy has not been measured, but the mechanism does not depend on dose.

The practical point is consistency. The prescription label's own instruction is that "Food may affect EPIDIOLEX levels" and that dosing should be consistent with respect to meals (FDA prescribing information). If you titrate while sometimes taking CBD with dinner and sometimes on an empty stomach, you are changing two variables at once and your log will be useless. Pick one pattern, usually with a meal that contains some fat, and keep it.

Can you take too much?

The prescription trials show what high doses do in a controlled population. At 10 and 20 mg/kg/day, somnolence occurred in 23% and 25% of patients versus 8% on placebo, decreased appetite in 16% and 22% versus 5%, and diarrhea in 9% and 20% versus 9% (FDA prescribing information). Those are 700 to 1,400 mg per day for a 70 kg adult, far above any consumer serving, and they show that the dose-response for unwanted effects is real.

Liver enzymes are the one laboratory signal worth knowing about. The label states plainly that the product "can cause dose-related elevations of liver transaminases," with elevations reported in 8% and 16% of patients at the two doses versus 3% on placebo, and the risk rises with concurrent valproate (FDA prescribing information). NCCIH lists the same category of effect, along with the plain observation that "CBD may have side effects, including decreases in alertness, changes in mood, decreased appetite" (NCCIH).

In a 2018 survey of 2,409 CBD users, 30.80% reported at least one adverse effect, and the authors summarized it as roughly "One out of every three users" (Corroon and Phillips, 2018). Most were minor. The lesson for dosing is simple: more is not better by default, the common unwanted effects are sleepiness and digestive upset, and the fix for both is to step back down to the last serving that produced neither.

How do you titrate over two weeks?

A worked example for a 50 mg/mL tincture, taken with the evening meal:

Days	Serving	Dropper	What to record
1 to 3	12.5 mg	0.25 mL	Time taken, any sleepiness, any stomach upset
4 to 7	25 mg	0.5 mL	Same, plus whether anything about your routine changed
8 to 10	37.5 mg	0.75 mL	Same
11 to 14	50 mg	1.0 mL	Same, then decide whether to hold or step back

Move to the next row only if the previous row produced nothing you disliked. If a row produces sleepiness during the day or loose stool, go back one row and stay there. Those two effects are the most frequently reported in the safety literature (Iffland and Grotenhermen, 2017), and they are the signal that you have passed the serving that suits you.

The prescription schedule follows the same logic at a different scale: 2.5 mg/kg twice daily for one week, then 5 mg/kg twice daily, with further increases weekly as tolerated (FDA prescribing information). Weekly steps work for a drug with monitoring. For a consumer product with a long half-life, 3 to 4 days per step is a reasonable compromise between patience and information.

Do not titrate past the point where you stop noticing a difference. The 1,793-person consumer trial saw its results at 15 mg per day (Saleska et al., 2023). If 25 mg and 50 mg feel identical in your log, 25 mg is your number and the second half of the bottle lasts twice as long.

How do you read strength on a label (1500 mg per 30 mL)?

Three numbers on a tincture label matter: total cannabinoid milligrams, volume, and serving size. Total milligrams is the headline figure and the one that makes bottles look comparable when they are not. A 1500 mg bottle at 30 mL (50 mg/mL) and a 3000 mg bottle at 60 mL (50 mg/mL) are the same strength; a 1500 mg bottle at 15 mL (100 mg/mL) is twice as strong per dropper. Always compute mg per mL before comparing prices or servings.

Serving size on the label is the manufacturer's suggestion, usually 0.5 mL or 1 mL. It is not a dose derived from your weight or the research. Regard the arithmetic as fixed and the serving as adjustable: the mg-per-mL number tells you what each dropper mark delivers, and the titration table above tells you which mark to use this week.

Gummies and softgels are easier: the label states milligrams per piece, and a 25 mg softgel is 25 mg regardless of bottle count. Still check the certificate of analysis. In the JAMA sample, oils were the most accurately labeled format at 45.00% and vaporization liquids the least at 12.50%, and THC was detected in 18 of 84 products at up to 6.43 mg/mL (Bonn-Miller et al., 2017). A product labeled "THC-free" needs a lab report that shows non-detect at a stated limit of quantitation, not a marketing line.

Do drug interactions matter?

The grapefruit comparison comes up because grapefruit juice blocks the same intestinal enzyme. FDA explains that many drugs are broken down with the help of "a vital enzyme called CYP3A4 in the small intestine," and that when grapefruit blocks it, "the juice lets more of the drug enter the blood" (FDA, Grapefruit Juice and Some Drugs Don't Mix). FDA's list of affected drug classes includes statins, some transplant drugs, some heart-rhythm drugs and some antihistamines. A reasonable rule of thumb, cited to that page, is that any prescription carrying a grapefruit warning deserves a pharmacist conversation before adding CBD.

The measured interactions on the prescription label are modest in both directions. Fluconazole, a strong CYP2C19 inhibitor, raised CBD exposure by 22% (AUC) and 24% (Cmax); rifampin, a strong CYP3A4 and CYP2C19 inducer, lowered it by 32% and 34%; itraconazole, a strong CYP3A4 inhibitor, changed CBD exposure by less than 10% (FDA prescribing information). The larger effect runs the other way: CBD raised the active metabolite of clobazam, a CYP2C19 substrate, approximately 3-fold.

The 2017 safety review cautions that many of the enzyme-inhibition findings come from test-tube work at CBD concentrations well above what a consumer serving produces, and that clinical interaction studies are still needed (Iffland and Grotenhermen, 2017). That is a reason for proportion, not for ignoring the question. As the Harvard pharmacist author put it, "CBD has the potential to interact with many other products" (Harvard Health). Bring the bottle, or the mg-per-mL number, to the pharmacist who already knows your prescriptions.

How do you log and adjust?

The log exists to defeat two biases. The first is expectation: the night you start, you will notice everything. The second is drift: by day 10, you will have forgotten what day 2 felt like. Writing one honest line per day fixes both. Rate the thing you care about on a 1 to 5 scale, the same scale every day, and write the number before you write any explanation.

Hold the serving steady while you log. In the 1,793-person consumer trial, sleep disturbance was scored weekly over a baseline week plus 4 weeks of use, and the within-group change was measured against each person's own starting point (Saleska et al., 2023). You can borrow the design: one week of baseline notes before you start costs nothing and makes the next two weeks interpretable.

Adjust in one direction at a time. If the log shows nothing after 4 days at a step, move up one row in the titration table. If it shows daytime sleepiness or digestive upset, the most common unwanted effects in the safety literature (Iffland and Grotenhermen, 2017), move down one row and stay. If two consecutive steps look the same, the lower one wins. And keep the pharmacist loop open: FDA notes that unapproved CBD products have not been evaluated for how they could interact with other drugs (FDA), so any new prescription is a reason to re-check.

This page is general information, not personalized guidance. FDA has not evaluated these statements.