Testosterone replacement is one molecule delivered six different ways, and the route changes almost everything about the experience except the drug. It changes how steady your levels are, how fast a dose error can be corrected, how much you pay, what safety warning applies, and what task you have to remember.

The pharmacology behind those differences is not complicated. Testosterone cypionate has a half-life of roughly eight days, so injection frequency determines the size of the peak-to-trough swing. Transdermal absorption varies several-fold between individuals, so gel dosing is empirical. Pellets release on a declining curve set at insertion, so the dose is fixed for months. Each of those facts produces a specific trade-off, and none of them makes one route correct for everyone.

The Verdict

Match the route to the constraint that binds hardest. Injections cost the least and allow the fastest dose correction, which matters most in a first year when the right dose is still being found. Gel and cream avoid needles and give steady daily levels, at a higher price and with a genuine transfer precaution around partners and children. Pellets remove the recurring task entirely, at the cost of reversibility — a dose that runs high cannot be reduced without surgical removal. Whichever route, monitoring is the same: levels, hematocrit, PSA and sensitive estradiol at 6–12 weeks, then every 6–12 months.

The six routes side by side

Prices are typical US self-pay ranges for medication and dispensing. They exclude consults and labs, which are usually the larger line. Level-stability descriptions assume the dosing frequency listed.

RouteHow it is givenLevel stabilityWhat it asks of youTypical drug costHow fast a dose error can be fixed
Injection — intramuscular or subcutaneous Testosterone cypionate or enanthate, weekly or split twice weekly Peak 24–48h after the dose, trough before the next; splitting the weekly dose flattens the swing One or two self-injections a week $30–$100/mo for the drug Dose changes take effect within a week
Injection — long-acting undecanoate Oil-based depot, dosed roughly every 10 weeks in clinic Very flat across the interval Four to five clinic visits a year Higher than generic cypionate; varies by payer A dose cannot be walked back for weeks
Transdermal gel or cream Applied daily to shoulders, upper arms or axilla Steady once absorption is established, but absorption varies widely between men Daily application plus a wash-and-cover routine $40–$300/mo depending on branded vs compounded Dose changes take effect within days
Subcutaneous pellets Implanted in the hip or flank through a small incision, every 3–5 months Supraphysiologic for the first weeks, then declining toward the end of the interval Three to four in-office procedures a year $500–$1,000 per insertion A dose cannot be reduced without surgical removal
Oral undecanoate capsules Twice daily with a fat-containing meal Peaks and troughs within each day; food-dependent Two doses a day, both with food $200–$600/mo Dose changes take effect within days
Nasal gel Applied to each nostril two to three times daily Short pulses rather than a steady level Two or three doses a day $300–$500/mo Dose changes take effect within days

Where injections lead

  • Cost. Generic cypionate or enanthate runs $30–$100 a month, the lowest of any route by a wide margin.
  • Reversibility. A dose that produces a high hematocrit or high estradiol can be reduced at the next injection, with the effect visible within one to two weeks.
  • Predictability. The pharmacokinetics are well characterised, so a level drawn at a consistent point in the interval is straightforward to act on.
  • Titration granularity. Dose can be adjusted in small increments, and frequency can be split further to flatten the curve without changing the weekly total.
  • Independence from absorption variability. The full dose is delivered regardless of skin type, body fat, showering habits or application technique.

The route choice does not settle where to inject. Four intramuscular labels name the gluteal muscle: Depo-Testosterone, generic enanthate, AZMIRO and AVEED. Subcutaneous Xyosted names the abdomen. No testosterone injection label names the thigh or the deltoid.

Where cream and gel lead

  • No needles. For men with genuine needle aversion, this is the difference between adhering to therapy and abandoning it.
  • Flatter daily curve. Daily application produces a smaller peak-to-trough swing than any injection interval, which tends to keep estradiol lower.
  • Fast washout. Stopping the drug clears it within days, which is useful if an adverse effect appears or a decision is being reconsidered.
  • Site and dose flexibility. Compounded creams allow fine dose adjustment, and scrotal application achieves target levels at lower doses in some men.
  • No procedure and no clinic visit. Refills are a pharmacy transaction; nothing about the route requires an appointment.

Secondary Transfer Evidence Across Topical Formulations

Secondary transfer of topical testosterone carries a boxed warning because direct skin contact can virilize children. The FDA mandated this warning in 2009 alongside formal risk mitigation requirements. Postmarket safety reports documented genital enlargement and advanced bone age. Some physical changes were not reversible. One report noted potential transfer from contaminated clothing, towels, or bed linens. The identical boxed warning on AndroGel, Testim and Fortesta states that children should avoid contact with unwashed or unclothed application sites, and that healthcare providers should advise patients to adhere strictly to the instructions for use.

Controlled transfer trials evaluated this risk. In an AndroGel study with 38 couples rubbing skin-to-skin for 15 minutes, every unprotected female partner reached a serum testosterone more than twice baseline at some point. Covering the treated skin with a shirt completely prevented transfer. Testim yielded very different results. In a 30-couple study, 15 minutes of skin contact increased female partner testosterone at least sixfold. A shirt failed to stop transfer when put on 15 minutes after application. Partner hormone levels still rose roughly threefold. A separate Testim trial showed long sleeves prevented transfer only after a one-to-four-hour delay. Products are not interchangeable. The AndroGel label states its dose and application site are not substitutable with other topical products.

Prescribing information defines different approved application sites for each branded gel. Approved sites for AndroGel 1% include the upper arms, shoulders, and abdomen. The label excludes the chest, back, and genitals. Testim and Vogelxo restrict application strictly to the upper arms and shoulders. Fortesta uses a 2% concentration and directs application to the front and inner thighs. Required washing intervals also differ. AndroGel directs patients to wait at least five hours before showering or swimming. Testim, Fortesta, and Vogelxo specify a two-hour window. Warm water and soap removed 99% of residual drug in Vogelxo hand-washing studies. Application-site washing reduced residual testosterone by 94.3% in Fortesta testing.

Patient factors like body fat are widely assumed to dictate transdermal absorption. The clinical trial data contradict that assumption. Prescribing information for AndroGel reports no observed differences in steady-state daily testosterone concentrations based on body mass index, age, or hypogonadism etiology. The manufacturer evaluated none of the other popular theories, such as skin thickness, hydration, or rubbing technique.

Interest in scrotal application stems from the high steroid permeability of thin genital skin. In a three-phase crossover study by Iyer and colleagues, scrotal cream produced rapid, dose-dependent testosterone peaks using lower doses than standard transdermal routes. A second Australian crossover study showed a 5% cream achieved bioequivalence to a 1% gel. Those pharmacokinetic data cannot be applied to American compounding. Both studies used regulated Australian products manufactured under strict protocols. No FDA-approved topical testosterone allows genital application. Beyond a single-patient case report, United States compounded creams lack published pharmacokinetic data. Transference risks from scrotal application have never been studied.

Where pellets lead

  • Adherence. There is no daily or weekly task to forget, which is the failure mode that ends more protocols than side effects do.
  • No transfer risk. Nothing sits on the skin, so partners and children are not a consideration.
  • Steady levels through the middle of the interval. Once past the early peak, the curve is flat for weeks at a time.
  • Discretion and travel. Nothing to carry, refrigerate, or explain at a border; no controlled-substance prescription to transport.
  • Fewer decisions. For a man who has already found a stable dose over a year or more on another route, the protocol becomes three or four appointments a year.

Pellet Extrusion Rates and Compounded Formulations

Pellet extrusion occurs in roughly one in ten procedures in prospective clinical trials. The official product label leaves that risk unquantified. In the United States, Drugs@FDA lists Testopel as the only approved testosterone pellet.

Its prescribing information notes post-marketing reports of extrusion, infection, cellulitis, and abscess at the implant site. Most of those complications appeared within the first month. Because these events are reported voluntarily, the manufacturer publishes no incidence rates.

Trial data provide specific frequencies. In a randomised trial of 251 procedures by Kelleher and colleagues, extrusion occurred in 12% of pellet-wash procedures and 11.1% of controls. Bleeding or bruising followed 8.8% of insertions, and infection developed in 4.0%.

Among patients who required antibiotics for an infection, 60% subsequently lost a pellet through extrusion. A 2002 trial across 400 procedures found extrusion after 11.2% of control insertions and 9.2% of antibiotic-soak insertions. The authors identified extrusion as the most common adverse effect of pellet therapy.

Published rates diverge sharply across study designs. A retrospective review by Conners and colleagues recorded extrusion in 7.5% of standard-technique procedures and 0.8% of modified insertions. In a 2023 study of transgender men by Cheng and Skolnick, 13.33% of patients experienced an extrusion.

Yet a 2021 retrospective by Donovitz reported extrusion in under 3% of male procedures across 1.2 million insertions. The difference stems from methodology and source material.

The prospective trials tracked every patient directly to log complications. The 1.2-million-procedure paper drew from clinic records across a single network dispensing compounded implants.

Anatomical site and clinician variation also alter trial outcomes. In a 2001 study by Kelleher and colleagues, extrusion reached 12% at the hip compared to 5% in the abdomen. The odds ratio was 2.6, though hormone absorption remained equivalent between both sites.

Their 2002 trial also demonstrated a statistically significant operator effect. One clinician had substantially more total and infection-related extrusions than colleagues. Differences in clinical experience or insertion technique did not explain that disparity.

Regulatory standards divide approved pellets from compounded alternatives. The Food and Drug Administration does not verify the safety, effectiveness, or quality of compounded drugs before marketing. Under federal law, 503B outsourcing facilities must follow current good manufacturing practice, but 503A compounding pharmacies do not.

A 2020 report from the National Academies of Sciences, Engineering, and Medicine recommended evaluating all hormone pellets for the difficult-to-compound list. The committee cited complex delivery mechanisms and absent bioavailability requirements. Manufacturing inspections have uncovered contamination risks.

An FDA warning letter dated 2 April 2025 cited an outsourcing facility for missing cleaning validation on pellet equipment and holding sterile products under insanitary conditions. Another warning letter dated 20 September 2021 cited distribution of unapproved pellet strengths, alongside unapproved testosterone and anastrozole combinations.

Medicare coverage reflects that distinction. Billing guidelines from the Centers for Medicare & Medicaid Services cover Testopel for approved indications up to six pellets, excluding unapproved products and pellet wastage.

What each route asks you to monitor

The core monitoring set does not change with the route. What changes is which risk is most likely to appear and when the blood draw should be timed.

Risk or effectWhich routes it applies toHow it is monitored
Erythrocytosis (rising hematocrit) Highest with injections, particularly at longer intervals and higher peaks Hematocrit at 3, 6 and 12 months, then annually; most prescribers act above roughly 54%
Secondary transfer to partners or children Specific to gels and creams; carries an FDA boxed warning for virilization in children Apply to covered skin, wash hands, allow drying time, cover the site with clothing
Site reaction, extrusion or infection Specific to pellets; extrusion occurs in a small but consistent share of insertions Wound checks after insertion; report drainage, swelling or a visible pellet
Blood pressure increase Boxed warning on oral undecanoate and on the weekly subcutaneous auto-injector Baseline and periodic blood-pressure monitoring; not a first choice in uncontrolled hypertension
Suppressed fertility All exogenous testosterone routes; nasal gel suppresses the axis least because of its short pulses Semen analysis if fertility matters; discuss hCG, a SERM, or sperm banking before starting
Estradiol elevation Follows peak height, so highest with widely spaced injections and early-interval pellets Sensitive LC/MS estradiol at 6–12 weeks; treat the pattern, not the number alone
Worsening of untreated sleep apnea All routes Screen for apnea before starting; treat it first if present
Prostate events All routes Baseline PSA, recheck at 3–12 months and annually; a rise over 1.4 ng/mL in a year warrants urology input

Two boxed warnings are route-specific and worth knowing before a first prescription. Transdermal gels carry an FDA boxed warning for secondary exposure, because skin contact can transfer enough testosterone to cause virilization in children. Oral testosterone undecanoate and the weekly subcutaneous auto-injector carry boxed warnings for blood-pressure increase, which makes them a poor first choice in uncontrolled hypertension.

Contraindications that apply to every route

None of these is a delivery-method question. They apply to testosterone therapy in any form, and screening for them belongs before the route is chosen.

  • Active breast cancer, or prostate cancer without specialist input
  • An unevaluated prostate nodule or an elevated baseline PSA
  • Baseline hematocrit above roughly 50%
  • Untreated severe obstructive sleep apnea
  • Uncontrolled heart failure, or a cardiovascular event within the past three to six months
  • Known thrombophilia or prior unprovoked venous thromboembolism
  • A desire to conceive in the near term — discuss hCG, a SERM, or sperm banking first

On cardiovascular safety, the TRAVERSE trial published in 2023 randomized roughly 5,200 middle-aged and older men with hypogonadism and cardiovascular risk to transdermal testosterone or placebo, and found no excess of major adverse cardiac events. It did find higher rates of pulmonary embolism, atrial fibrillation and acute kidney injury on testosterone. That evidence applies to the drug, not to any particular route.

February 2025 Class-Wide FDA Labelling Changes

On 28 February 2025, the FDA informed the sponsors of testosterone products of class-wide labelling updates following its review of the TRAVERSE trial and the results of required postmarket ambulatory blood pressure monitoring studies. Across every approved testosterone product, the FDA:

  • removed the Boxed Warning language about an increased risk of adverse cardiovascular outcomes
  • added the TRAVERSE trial results to every testosterone product's labelling
  • retained the Limitation of Use statement for age-related hypogonadism

The FDA also required product-specific blood-pressure data where studies finished, and added a new warning about increased blood pressure to testosterone products that lacked that labelling. The FDA stated that premarket ambulatory blood pressure monitoring of two products had raised concerns that testosterone could increase blood pressure. A blood-pressure boxed warning already sits on oral testosterone undecanoate and on the weekly subcutaneous auto-injector.

Removal of the cardiovascular language from the Boxed Warning applies class-wide, so it does not favour one delivery route over another. Blood-pressure labelling now reaches products across the class rather than sitting with two of them. The boxed warning still applies only to oral undecanoate and the weekly subcutaneous auto-injector, which is why those two stay a poor first choice in uncontrolled hypertension. The oral capsule therefore carries a blood-pressure warning rather than a liver warning. The liver-enzyme question on oral testosterone traces that reputation back to a different class of drug.

The update does not alter the secondary-transfer boxed warning on transdermal gels and creams, and erythrocytosis risk remains highest with injections. If you are choosing a route today, this regulatory update changes the warning text on your prescription label while leaving the necessary clinical monitoring schedule in place. A compounded preparation has no FDA-reviewed label, so this class-wide change does not reach it the same way. We examine those distinctions in our review of compounded vs brand-name testosterone. Our guide to TRT side effects and risks outlines cardiovascular outcomes and hematocrit tracking in detail.

How to decide

  1. First year, dose not yet settled. Favour a route you can adjust quickly — split injections or daily transdermal — so a lab finding at week eight can be acted on rather than waited out.
  2. Young children at home, or a partner who is or may become pregnant. Transfer risk is a real consideration; injections or pellets remove it.
  3. Needle aversion that has not resolved. Transdermal is the route that keeps men on therapy; a compounded cream also allows finer dose steps than a standard gel pump.
  4. Adherence is the repeated failure point. Pellets remove the task, and are a more reasonable choice once a stable dose is already known from another route.
  5. Fertility is a live concern. Consider whether testosterone is the right drug at all. Enclomiphene, clomiphene or hCG raise endogenous production without suppressing the axis, and nasal gel suppresses it least among testosterone routes.
  6. Uncontrolled blood pressure. Avoid oral undecanoate and the weekly auto-injector until it is managed.

Frequently Asked Questions

Which TRT delivery method should I choose?

It depends on which constraint binds hardest. If cost and dose control matter most, injections are the cheapest route and the easiest to titrate. If needles are not acceptable, daily gel or cream gives comparable levels with a transfer precaution to manage. If a daily or weekly task is the thing you will not sustain, pellets remove it entirely at the cost of reversibility and price. If fertility is a live concern, nasal gel suppresses the axis least, and a SERM or hCG-based approach may be a better conversation than testosterone at all.

Why do twice-weekly injections get recommended over weekly?

Testosterone cypionate has a half-life of roughly eight days, so a single weekly dose produces a peak within a day or two and a trough before the next injection. Splitting the same weekly total into two smaller doses roughly halves that swing. Men who report a mood and energy decline in the last two days of the interval frequently stop reporting it after the split. It is the same drug and the same monthly cost, with two injections instead of one.

Are pellets worth the cost?

For a man who has tried other routes and cannot sustain them, the convenience is real and the levels are steady across most of the interval. Two things make them a poor default for a first year. Levels run supraphysiologic in the first weeks and decline toward the end, which is the opposite of what titration wants. And a dose that turns out to be too high cannot be reduced without surgical removal, so a hematocrit or estradiol problem at week six has to be managed around rather than corrected.

How risky is testosterone gel around children?

Serious enough that the FDA requires a boxed warning. Skin-to-skin contact can transfer enough testosterone to cause virilization in children — early pubic hair, genital enlargement, advanced bone age — and some cases have not fully reversed. The precautions are practical rather than onerous: apply to skin that clothing will cover, wash hands immediately, let the site dry before dressing, and avoid contact with the site. Families with young children often choose injections for this reason alone.

Can I switch methods later?

Yes. Delivery route is not a permanent decision, and switching is routine. Many men start on gel and move to injections after seeing the cost difference, or start on injections and move to gel if injection anxiety does not settle. Expect a fresh level check 6–12 weeks after any switch, timed consistently relative to the dose, because the peak-to-trough shape changes with the route.

Does the delivery method change the monitoring schedule?

The schedule is broadly the same — levels, hematocrit, PSA and sensitive estradiol at 6–12 weeks after starting or changing, then every 6–12 months. What changes is the timing of the draw. On injections, draw at a consistent point in the interval, usually mid-interval or at trough, and record which. On gel, draw 2–8 hours after application. On pellets, a level at week 4 and again near the end of the interval shows the actual range rather than a single point on a declining curve.

Which method is best for keeping estradiol in range?

Aromatization follows peak testosterone concentration, so routes with lower peaks tend to produce lower estradiol. Split subcutaneous injections and daily transdermal application both keep peaks modest. Widely spaced injections and the first weeks after a pellet insertion produce the highest peaks and the highest estradiol. This is a reason to fix the delivery pattern before reaching for an aromatase inhibitor, which is over-prescribed relative to the evidence and carries its own risks at low estradiol.

Related