Fueling a Long Indoor Rowing Session

The context

Two hours on the erg is not a hard session that happens to last a long time. It is a session that spends more carbohydrate and more fluid than it can replace, and the difference between a good one and a flat one is usually decided in the last forty minutes.

The first hour is fine. The rate is there, the grip is still shut on the handle, and the lats are pulling properly through the catch. Then something changes: the hands open on the handle, the drive gets shorter, the strokes go ragged at the finish, and the last forty minutes feel nothing like the first forty. That change is not a fitness failure, and it is not the damper either.

For endurance exercise of half an hour or more, the two things most likely to be wearing you down are dehydration and carbohydrate depletion, with gastrointestinal problems, hyperthermia and hyponatraemia behind them and the more serious the longer the event runs ([2] Jeukendrup AE 2011). The rowing-specific arithmetic points the same way. Indirect calorimetry on 14 elite open-class male rowers put exercise energy expenditure between 15.6 and 49.8 kcal a minute across three intensity zones, with carbohydrate supplying between 46.4% and 100% of the fuel, and the authors conclude that carbohydrates are dominantly utilised even during low-intensity rowing training ([1] Winkert 2022).

That last figure is the one to sit with. A long indoor row is not a flat two hours of one thing: carbohydrate is already most of the bill before the hard part.

The core idea

Three budgets running on three different clocks: carbohydrate in the tank, fluid leaving the body, and the glycogen the session started with.

The glycogen clock

This is the one that explains the last third of the piece.

Seven endurance-trained cyclists rode at 71% of maximal oxygen consumption until they were spent. Unfed, they stopped at 3.02 hours, their blood glucose having fallen to 2.5 ± 0.5 mmol/L, with muscle glycogen draining at 51.5 mmol of glucosyl units per kilogram per hour through the first two hours and, at a slower rate, 23.0 in the third and final hour ([3] Coyle 1986).

Fed a glucose polymer drink, the same riders went an extra hour, to 4.02 hours, with blood glucose held between 4.2 and 5.2 mmol/L. That extra hour cost 5 mmol of glucosyl units per kilogram per hour, a fraction of the earlier rate ([3] Coyle 1986).

That is the shape of a long session in two numbers: the opening part is paid out of the tank you arrived with, and the part where it stops feeling like rowing is paid out of a tank you did not refill. These were cyclists rather than rowers, at a steady 71% of maximum, so the direction transfers but the exact minute does not.

The carbohydrate clock

What you eat in the hours and days before a session sets how much carbohydrate is available to it, and refuelling between sessions is part of the same arithmetic rather than a separate question ([4] Burke 2011). The store itself is built by what you eat day to day: endogenous glycogen stores are maximised by a high-carbohydrate diet of 8-12 g of carbohydrate per kilogram of body mass per day ([6] Kerksick 2017).

For the meal on the day, the guideline is 1-4 g per kilogram, and the practice record is ordinary. Among 50 athletes who completed endurance events of 2.5 hours or more, 80% met the pre-competition meal guideline in practice, and the 38% who correctly identified that guideline were eating 1.4 ± 0.6 g/kg ([7] Sampson 2024).

The rate during the piece is where the numbers get specific. Thirty to 60 g of carbohydrate an hour is the appropriate target for sport of longer duration, and events over about 2.5 hours may benefit from intakes of up to 90 g/h ([4] Burke 2011). The ceiling is physiological rather than arbitrary: a single carbohydrate source can be oxidised at rates up to approximately 60 g/h, and that is the recommendation for exercise lasting 2-3 hours, while higher rates require a blend of multiple transportable carbohydrates rather than a larger gulp of the same one ([5] Jeukendrup A 2014). Two caveats come from the same source. The intake has to be scaled down when the absolute exercise intensity is low, because oxidation rates are low then too ([5] Jeukendrup A 2014). And the advice is independent of body weight and training status ([5] Jeukendrup A 2014).

Delivery matters as much as the total. For extended bouts of more than 60 minutes at high intensity, above 70% of VO2max, carbohydrate should be consumed at about 30-60 g/h in a 6-8% carbohydrate-electrolyte solution, 6-12 fluid ounces, every 10-15 minutes through the whole bout, and this matters most once the bout runs past 70 minutes ([6] Kerksick 2017). A 6-8% drink is 60-80 g of carbohydrate a litre, so a two-hour piece needs between three-quarters of a litre and two litres to cover a 30-60 g/h target.

The fluid clock

The goal of drinking during exercise is to prevent excessive dehydration, more than 2% of body weight lost, and excessive changes in electrolyte balance that would compromise performance. Because sweating rates and sweat electrolyte content vary considerably between individuals, customised fluid replacement programmes are recommended, and an individual's sweat rate can be estimated by measuring body weight before and after exercise ([9] American College 2007). That estimate is for one session and one set of conditions. During exercise, a drink carrying electrolytes and carbohydrates can give you something water alone does not, under certain circumstances ([9] American College 2007).

How much leaves the body is not trivial in a warm room. Daily water losses of 4-10 L and sodium losses of 3500-7000 mg are possible in active athletes in hot weather, and both water and sodium have to be replaced to get back to normal total body water ([10] Shirreffs 2011).

The protein clock

Protein works on a daily clock rather than a post-session one. A consensus opinion puts protein intakes at 1.3-1.8 g/kg/day, taken as 3-4 isonitrogenous meals, as the range that maximises muscle protein synthesis, with experienced athletes needing less and high-frequency or high-intensity training needing more ([8] Phillips 2011).

For the session itself the window is short and specific: high-quality protein from immediately after to two hours after the session stimulates robust increases in muscle protein synthesis, and a 20-40 g dose, 0.25-0.40 g/kg of body mass per dose, every three to four hours does the same ([6] Kerksick 2017).

The how-to

For the two-hour Saturday.

Before. Eat the food you have eaten a hundred times before, carbohydrate-rich and low in fibre, and arrive with the tank full rather than empty. 8-12 g/kg/day of carbohydrate maximises the store ([6] Kerksick 2017). The meal on the day is 1-4 g/kg ([7] Sampson 2024), eaten early enough to have been digested before you start.

During. 30-60 g of carbohydrate an hour ([4] Burke 2011), delivered as a 6-8% drink sipped every 10-15 minutes through the bout. That rate belongs to extended bouts of high intensity, above 70% of VO2max ([6] Kerksick 2017). Build the habit, because the habit is what fails rather than the intention: in the survey of 50 endurance athletes, only 36% met the during-competition guideline of 60-90 g/h, and the athletes who correctly identified that guideline still ate 56 ± 20 g/h, with no relationship between knowing the number and eating it ([7] Sampson 2024).

The other habit that fails is drinking to a schedule rather than to your own sweat rate. Weigh in before and after a session and you have an estimate of what you lost that day ([9] American College 2007): the gap covers what you drank and passed as well as what you sweated. Use it to start from, then drink to thirst, rather than forcing fluid down to hit a number ([11] Altieri 2026).

Watch the damper too. It sets the drag factor, which decides how much air every stroke moves, so raising it for the second half makes every stroke cost more at the point you can least afford it.

Practise on a shorter session first. Take the same drink at the same rate on a 60-minute piece, then a 90-minute one, before the two-hour one. Knowing what to do does not make you do it ([7] Sampson 2024), and the only reliable way to find out whether you can hold 30-60 g an hour without your gut turning over is to find out where failing costs you nothing.

After. If another session is inside four hours, the recovery window is not optional: aggressive carbohydrate refeeding at 1.2 g/kg/h, or 0.8 g/kg/h of carbohydrate combined with 0.2-0.4 g/kg/h of protein, restores glycogen quickly, with a preference for higher-glycaemic carbohydrate sources ([6] Kerksick 2017). A 20-40 g protein dose in the first two hours after the session ([6] Kerksick 2017). And 1.3-1.8 g/kg across the day as 3-4 meals ([8] Phillips 2011).

The trade-offs

Fuel against gut. The ceiling on carbohydrate is a tolerance ceiling as much as a physiological one. Gastrointestinal problems in long events occur frequently, and they seem to be highly individual, perhaps genetically determined, and related to the intake of highly concentrated carbohydrate solutions, hyperosmotic drinks, and of fibre, fat and protein ([2] Jeukendrup AE 2011). That is the trade-off the 90 g/h figure hides: pushing the rate up means a more concentrated drink, and the drink is often what causes the problem. Start at 30 g/h and work up over a few sessions.

A plan against thirst. A long session can also go wrong the other way. Exercise-associated hyponatraemia is multifactorial, driven by excessive fluid intake and by non-osmotic vasopressin release induced by the exercise itself, with sodium loss through sweat playing a less important role than the volume you drink. It presents as nausea, vomiting, headache and confusion, and it has to be told apart from hypoglycaemia, orthostatic hypotension, vasovagal syncope and heat stroke ([11] Altieri 2026). The same review is blunt about the fix: recommendations to drink in anticipation of thirst during physical exertion should be replaced by a strategy of drinking when thirsty ([11] Altieri 2026).

The advice earlier and this pull against each other, so it is worth saying which wins: thirst. The scale tells you what one session cost you; thirst tells you what to drink during the next one. Both appear in the same review of endurance fuelling, which reports hyponatraemia occasionally in slower competitors with very high intakes of water or other low-sodium drinks ([2] Jeukendrup AE 2011).

Where coaching stops. If you use insulin, the carbohydrate you take during a session is part of your glucose management rather than a coaching choice, and the plan has to be written with your prescriber. A history of hyponatraemia during exercise, kidney or heart disease, or fuelling for a session in extreme heat all belong with a clinician and a registered dietitian. The formal position from the joint nutrition and sports-medicine bodies is that athletes should be referred to a registered dietitian nutritionist for a personalised nutrition plan ([12] Thomas 2016).

Stop rather than fuelling through if nausea, vomiting, headache or confusion appears during a session, or if the symptoms look more like a medical problem than low fuel ([11] Altieri 2026).

What to do tomorrow

• Weigh in before and after your next long session to estimate your own sweat rate, then drink to thirst rather than to a number ([9] American College 2007). ([11] Altieri 2026).
• Put a 6-8% carbohydrate drink within reach and drink from it every 10-15 minutes for the whole piece, if the piece is an extended bout of high intensity ([6] Kerksick 2017).
• Measure what you eat before a session against 1-4 g/kg ([7] Sampson 2024), then do it again on a shorter session until it is automatic.
• If you have never fuelled through a long session, start with a 60-minute piece at 30 g/h and build up from there ([4] Burke 2011).
• Decide the food for the moment you stop before you start, not later in the day ([6] Kerksick 2017).