How to pick your starting weight when you've never lifted

TL;DR: Start day one at roughly half of what you think you could do. Rate every set on a simple effort scale, then let the next session's weight adjust automatically. Within two or three sessions, you'll be within 5% of your true working weight. No spreadsheets, no guessing, no injuries.

The question you're really asking isn't "how much should I lift?" It's "how do I not embarrass myself and also not get hurt?" Those are good instincts, but they push you toward two bad answers: pick a number that sounds respectable, or pick a number so light it doesn't count.

Your first session isn't a commitment. It's a starting signal: a single data point the rest of your program uses to calibrate itself. You're not declaring your identity as a lifter. You're giving the system a number to work with so it can find the right one for you. That right number is discoverable, not guessable, and definitely not something a stranger on the internet can tell you from your height and bodyweight alone.

The two ways beginners get this wrong

There are only two real failure modes, and they're mirror images of each other.

Failure mode 1: You start too heavy. This is the more common one, especially for people with some athletic background or who've watched a lot of lifting content. You grab a weight that looks right (what a friend uses, or what a video said was "standard for a man your size") and muscle through eight reps. The first three feel fine. The fourth is uglier. By rep six, your back is rounding, your elbows are flaring, your knees are caving, or all three. You grind out the last two on willpower and walk away thinking "I did it."

The next morning, soreness is so bad you can't go up stairs. The session that was supposed to be 48 hours later gets pushed to 72, then 96, then next week. You miss it. The one after that, you show up but you're protective of whatever hurt (a shoulder, a lower back, a knee) and the session is half-effort. Within three weeks you've quit. The reason isn't laziness. It's that going too heavy wrote a check your recovery couldn't cash, and the interest compounds into soreness, dread, and injury.

Failure mode 2: You start too light and never adjust. This one is sneakier because it doesn't hurt. You pick the lightest dumbbells, do three sets of 12, feel nothing, and go home. Same weight next session. Same the session after. Six weeks in, you're no stronger, look the same in the mirror, and decide lifting "doesn't work for you." It wasn't the lifting. Your body adapts to the stimulus you give it, and a stimulus that never increases produces no adaptation. Light forever is not safe. It's invisible quitting on a slower timer.

Both failures come from the same root cause: treating your starting weight as a fixed number instead of a moving estimate. The fix is a protocol that expects to be wrong on day one and gets less wrong every session.

The 50% rule for day one

Here's the rule: whatever weight you think you can do for 8 reps, pick 50% of it for session one.

If you think you could bench the 45 lb bar plus a 25 on each side for 8, bench just the bar. If you think you could squat 135, squat 65. If you think you can do a regular push-up for 8, do incline push-ups against a counter instead. If you think you can curl the 25s, pick up the 12s.

Yes, it will feel stupid. Yes, it will feel like you're "not really working out." You will want to add weight mid-set. Don't.

Here's what you're actually doing with that "stupidly light" day one:

• Learning the movement pattern without distraction. You pay attention to which joints are moving, what your back is doing, where your weight sits on your feet.
• Mapping where you'll be sore tomorrow. Even 50% gives you some DOMS in muscles you've never asked to work through a full range. That tells your program which muscles actually did the work versus which ones compensated.
• Calibrating the RPE scale (next section). You can't rate effort honestly if every set is a 10. A 50% set gives you a reference for what a 5 feels like, so later you actually know the difference between a 7 and a 9.
• Banking a repeatable session. The single most predictive thing for whether a beginner is still training in six weeks is whether their first session was repeatable: not impressive, repeatable. The weight you walk away from feeling good is the weight you come back for.

You will out-earn day one within a single week. There's no scenario where starting at 50% meaningfully slows your long-term progress, and many where starting at 80% derails it.

Use RPE to find your weight in 3 sessions

RPE stands for Rate of Perceived Exertion. It's a 1-to-10 scale for how hard a set felt. For beginners, you don't need the whole scale. You only need three buckets, which is why we sometimes call this the "three-word" method.

After every working set, ask yourself one question: could I have done more reps with good form?

• "I missed reps" (you didn't hit the target rep count with clean form): the weight was too heavy. In RPE terms, this was a 10 or above. Drop weight next session.
• "I hit it exactly" (you got the reps, but the last one or two were genuinely hard, and a third would have been ugly): the weight was right. This is roughly RPE 7–8. Keep the weight next session and aim for one more rep, or add the smallest increment and repeat the rep target.
• "I beat it" (I finished the set and felt like I had 3+ clean reps left in the tank): the weight was too light. This is RPE 5 or below. Add a meaningful jump next session: 5 lb or more for upper body, 10 lb or more for lower body.

That's the whole protocol. Three buckets. One question. Now watch what happens across three sessions.

Session 1: You start at 50%. Almost every set is "I beat it." The next session's weights get a big bump.

Session 2: You land in the middle. A lift or two still feels light ("beat it"), one or two land in "hit it exactly." You now have real info: which lifts are you actually strong on, and which are weaker than you assumed? This is the first session where the program really starts to know you.

Session 3: Most lifts are "hit it exactly." A couple might be "missed" (those correct down). A couple still "beat it" (those keep climbing). By the end of session three, you are within 5% of your honest, sustainable working weight for every movement. Nothing you guessed beforehand could have gotten you closer.

RPE works because it replaces a question you can't answer ("what's the right weight?") with one you can ("did that feel hard?"). Your nervous system is a far better judge of load than any calculator, and it gives you that judgment for free after every set. You just have to listen. For more on why this self-rating approach is the backbone of sustainable training, see our post on RPE autoregulation.

Bodyweight exercises: what "start light" means

"Start at 50%" sounds obvious with dumbbells. But what if your program has push-ups, pull-ups, or dips, exercises where there's no weight to cut in half? The principle is the same; you just change the variable.

For push-ups, the progression is:

1. Wall push-ups. Stand arm's length from a wall, hands flat, push back and forth. This is push-up "empty bar": almost no load, but the exact same movement pattern.
2. Incline push-ups. Hands on a counter, a sturdy table, or a bench. The steeper the angle, the easier the rep. Lower the surface over time.
3. Knee push-ups on the floor. Same plank line from knees to shoulders: not sagging, not piking.
4. Full push-ups on the floor.
5. Decline or weighted push-ups once full push-ups for 10 reps are easy.

For pull-ups, start with dead hangs (just hold the bar), then assisted pull-ups (bands or an assisted machine), then negatives (jump up, lower slowly), then unassisted reps.

For dips, start with bench dips (feet on the floor, hands behind you on a bench), then parallel bar dips with a band, then unassisted.

The logic is identical to the 50% rule: pick the regression that makes rep one feel easy and rep eight feel like maybe two more were in the tank. If you can't find that zone, the movement is still too hard and you regress one more step. There is always a regression. "I can't do a push-up" is never a dead end: wall push-ups exist, and they work.

The one warm-up set rule

Before every working set at your target weight, do one warm-up set of the same movement at roughly 50% of the working weight for 5 easy reps. Not five sets. Not a full warm-up protocol. One set.

This does three things, all load-bearing:

1. It tells your joints (shoulders, hips, knees) what's about to happen. Cold tissues are stiffer and more injury-prone than warm ones.
2. It lets you rehearse the pattern with an easy load, so your first heavy rep isn't your first rep of the day. You catch any weirdness while the stakes are low.
3. It's a final sanity check on the weight. If the warm-up feels unexpectedly hard, that's useful intel before you add plates.

One set is the minimum viable warm-up. Beginners usually don't need more, because their working weights aren't heavy enough to demand a five-set ramp. More on when and how to add ramping in the warm-up sets protocol.

What Grov does for you

Everything above is a protocol you could run yourself with a notebook, and plenty of people do. What Grov adds is that it runs the protocol for you, so you don't have to remember the rules or do the arithmetic mid-workout.

When you onboard, Grov asks for a few things: your bodyweight, your training history (which, for most readers of this post, is "none"), and your self-rated fitness level. From those, it picks a starting weight for every lift in your program that sits at or below the 50% line, deliberately underloaded, because underloading is the move.

Then, after every set, you tap one of three buttons: "missed," "exactly," or "beat it." That's the RPE signal. Grov takes that signal and writes your next session's weights automatically. By session three, the weights in your plan reflect you, not the average beginner, because you taught the plan who you are, one set at a time.

This is the mechanism we believe in and why we built the product this way. The full version of the argument, why auto-regulated load beats fixed programs and why underloading day one is the fastest path to a program that fits you, is laid out in our thesis.

---

You don't need to pick the right starting weight. You need to pick a starting weight you can learn from. Then you need to show up for session two, and session three, and let the protocol do its work.

Start training →