Collagen for tendons is dosed at 15 g, three 5 g scoops, 30 to 60 minutes before training, with 50 to 100 mg of vitamin C. From Shaw, Am J Clin Nutr 2017: eight healthy men, a blood marker, not a tendon. At 15 g, TARE’s 350 g tub holds 23 doses, not the 35 its panel declares.
What follows is the working: where the 15 g came from, what it has and has not been shown to do, and what three scoops a session does to a tub whose own panel counts 35 servings of 10 g.
The last part is the one a seller has no reason to print about its own container. The tub lasts fewer uses than its own serving count implies. That is arithmetic rather than a defect, and it belongs on this page because the reader doing the division at the checkout should not be the first person to have done it.
How much collagen for tendons, and when?
Fifteen grams as a single dose, 30 to 60 minutes before the session that loads the tendon, on training days. That is three 5 g scoops. Take 50 to 100 mg of vitamin C with it, in the same drink or alongside it. There is no rest-day dose in this protocol, because every result quoted below was measured with loading rather than with powder alone.
Two of those instructions are the ones that get dropped. The vitamin C is part of the dose, not a nicety — it was in the arm of the study that produced the result, and the protocol figure is 50 to 100 mg taken with the dose. And the window is before, not after. The protocol page tags the same instruction T–45, which is the middle of that window rather than another number to remember.
The vitamin C is not in the tub. Under Load is one ingredient, collagen peptides, and nothing has been added to it — which means the cofactor is something a reader supplies separately rather than something already paid for. It is also unflavoured, so the drink it goes into is whichever one was already going to the session. That is the whole of the compliance argument for it: nothing in the day has to move.
Where does the 15 g figure come from?
One trial. Shaw, Lee-Barthel, Ross, Wang and Baar, Am J Clin Nutr 2017;105(1):136–143, compared 5 g against 15 g of gelatin taken with vitamin C one hour before loading. The 15 g arm roughly doubled serum PINP and the 5 g arm landed in between. The sample was n = 8 healthy males, and PINP is a marker measured in blood, not a tendon.
Read that twice, because the half usually reproduced is the first one. Eight men, and an endpoint drawn from blood. A rise in a synthesis marker is a reason to expect something; it is not a measurement of a tendon, and no amount of repetition upgrades it into one.
Tendon tissue itself appears in the wider literature rather than in that trial: cross-sectional area and stiffness increase at 15 to 30 g a day with loading — systematic review, JFMK 2026. The two words carrying that sentence are with loading. The powder is the smaller half of the intervention and the training is the larger one, which is also why the dose here is tied to a session rather than to a day.
The trials disagree, and the largest one that says no was run at this page’s own dose. Balshaw and colleagues, Medicine & Science in Sports & Exercise 2023 — n = 39, 15 g a day, 15 weeks of resistance training, the tendon measured by MRI rather than ultrasound — found no difference from placebo on any tendon measure. That trial is larger than most of the favourable ones, longer than most, better instrumented than most, and it was not funded by industry, which is the opposite of how the positive tendon results skew. It belongs on the same page as the review above rather than three clicks away from it.
So the honest summary of the dose is two sentences long. 15 g is the best-supported number available. The evidence underneath it is thinner than the confidence with which it is usually quoted. Both are true at once, and a brand that gives only the first sentence has told half of it.
Is more than 15 g better?
No. Across that range 30 g produced nothing beyond 15 g. More is not better, and doubling the scoop count doubles the cost of a session and nothing else.
The clearest negative result sits one tissue over, and it is quoted here for exactly the reason it is usually left out. Aussieker and van Loon, MSSE 2023 — 30 g after exercise produced no increase in muscle connective protein synthesis. That is a study of muscle, and the finding is null.
Collagen is not a muscle-building protein and TARE does not sell it as one. It is a pre-training dose with a stated dose-response and a stated ceiling, and the ceiling is the part of the response curve a seller has no commercial reason to mention. It is mentioned here because a reader who scales the scoop up is spending money to reach a plateau that was already reached at three scoops.
How many doses is a 350 g tub at 15 g?
Twenty-three. 350 ÷ 15 leaves 23 whole doses and one orphan 5 g scoop that cannot make a 24th. The Supplement Facts panel on the same tub declares 35 servings, because a declared serving there is 10 g, two scoops, not three. Both numbers are correct and they are counting different things.
| What is being counted | The figure |
|---|---|
| Declared serving, Supplement Facts panel | 10 g · two scoops · 35 servings |
| Protocol dose, this page | 15 g · three scoops · training days |
| Whole doses in a 350 g tub | 23, plus one 5 g scoop over |
| At four sessions a week | 40 days |
| At five sessions a week | 32 days |
The two duration lines are 23 ÷ 4 × 7 = 40.25 and 23 ÷ 5 × 7 = 32.2, written down as 40 days and 32 days. A duration is never rounded up here. An earlier draft of this arithmetic carried about 41 days and about 33; both were wrong, and both were wrong in TARE’s favour.
Which leaves the sentence this page exists to carry: at the dose recommended above, the tub is 23 uses, not 35. A serving count answers one question — how many 10 g servings are in the container. It has never answered the question a buyer is actually asking, which is how many training days the container covers, and reading the first as an answer to the second is the ordinary way to be surprised by a re-order date.
Why the label says 10 g and this page says 15 g
Because they are two different kinds of statement. The Supplement Facts panel is a serving declaration: 10 g, two 5 g scoops, 35 servings, and it is locked. The 15 g on this page is a directions statement — three scoops, 30 to 60 minutes before training — which describes how to use the tub rather than what the tub declares. The label states what is inside it. The site states how it is used.
That distinction is not a technicality invented to explain an inconsistency; it is the reason no front-panel figure anywhere on this container contradicts its own panel. A serving declaration is a fixed unit a panel is built on, and every derived figure on that panel — calories, protein, the servings count — is computed against it. A directions line sits outside that arithmetic and cannot move it.
It is also the single most portable thing on this page. Carried onto anyone else’s container, it turns the serving-size line into the first thing read rather than the last: how to read a supplement label is the long version, and the serving size is where nearly every surprise on a panel begins.
What collagen has not been shown to do
Two things belong here in plain words. Collagen has not been shown to add strength. The 2026 JFMK systematic review grades that evidence strong — GRADE A — against an effect on muscle strength, and the 2024 Sports Medicine meta-analysis that did find an effect measured it at SMD 0.19, trivially small and rated low-certainty by its own authors. And no randomised controlled trial shows reduced injury rates — no trial has reported one, at any dose. Any page implying otherwise, including this one, would be inventing it.
The second of those is the claim the category leans on hardest and it is the one with nothing under it. A marker that moves and a tissue property that changes with loading are both real findings, and neither of them is a count of injuries in a training population. No trial has reported that result, so nobody may imply it, including in the softened forms — "for resilience", "for staying available" — which mean the same thing while being harder to check.
The fair criticisms, in our own words
These are the objections a critic would raise, and there is no version of them TARE can argue with. Collagen carries zero tryptophan. It is low in leucine. Its DIAAS is effectively zero, which is the formal way of recording that it is a poor protein for the purpose most protein powders are bought for. And "it is hydrolysed gelatin" is largely true.
None of that is answered by a better sentence. It is answered by not selling collagen as a protein supplement, which is what the dose, the window and the ceiling above already describe. A pre-training dose with an honest limit is a different product from a protein powder, and the amino-acid profile is only an indictment of the second one.
The one protein figure that exists on paper is worth stating exactly, because it is the sort of number that grows in transit: 9.8 g measured per 10 g serving; 1.5 servings is 14.7 g by arithmetic. The 9.8 g is a laboratory result on lot 121125450 against a 9 g specification, by ISO 14891:2002. The 14.7 g is multiplication. Only one of those two numbers was ever measured, and it is not the larger one.
Where our scoop sits
Under Load is collagen Type I and III, bovine, one ingredient, in a 350 g tub with a 5 g scoop. The panel declares 10 g and 35 servings. This page asks for three scoops before training, which is 23 doses. Both of those sentences are printed on purpose and neither one is walked back by the other.
The certificate behind the tub is a real test of a real 350 g lot, and it is not an independent one. ViaCore Nutrition, El Reno, Oklahoma, is the manufacturer’s own laboratory — lot 121125450, analysis dated 12/22/2025. That is a weaker check than an outside laboratory, and a reader should hear it here rather than deduce it. What the document does and does not cover is transcribed row by row on Testing & Certification.
More in this cluster: what a branded collagen peptide is, and what it is not, whether collagen and creatine can be taken together, what a certificate of analysis proves, and what ours do not, and how to read a supplement label. Or read the documents themselves on Testing & Certification.