What is a stop-limit order?
Most order-type guides treat “stop-limit” as one thing with one number. It is two numbers, and the distance between them is the whole decision. Set them badly and you get the worst outcome the order can produce: the stop fires, nothing fills, and you are still holding the trade you meant to be out of.
Stop price and limit price — what is the difference?
The stop price is a condition. It does nothing except watch. When the market reaches it, the exchange takes an order it has been holding in reserve and places it into the order book. Until that moment nothing of yours exists in the market — no other trader can see it, and it cannot fill.
The limit price is the order that gets placed. It is an ordinary limit order, with all the properties of one: it fills at its price or better, and it does not fill at all if nobody meets it. Setting it is a promise to yourself — I will exit at this price, but not worse than this price.
Two different jobs, and confusing them is the most common beginner error with this order type. A useful reframe: the stop price answers when do I act? The limit price answers what is the worst I will accept once I act? A plain stop-loss market order answers only the first question and hands the second to the market.
How far apart should the two prices be?
This is the only real decision, so it is worth doing with numbers rather than instinct. Take a long position of 0.5 BTC opened at $60,000, with a stop price at $58,000. If the order filled exactly at the stop, the loss would be 0.5 × $2,000 = $1,000. Now choose where to put the limit price underneath it.
| Limit price | Offset below stop | Worst fill if it fills | Loss at that fill | Fails to fill when… |
|---|---|---|---|---|
| $57,990 | $10 (0.02%) | $57,990 | $1,005 | price drops more than $10 through the stop |
| $57,900 | $100 (0.17%) | $57,900 | $1,050 | …more than $100 |
| $57,600 | $400 (0.69%) | $57,600 | $1,200 | …more than $400 |
| $57,000 | $1,000 (1.72%) | $57,000 | $1,500 | …more than $1,000 |
| none (stop-market) | — | whatever the book holds | not capped | never — it always fills |
Our own worked example, not exchange data. Offsets are % of the stop price; loss figures assume the worst permitted fill and exclude fees. Reproduce any row as 0.5 × ($60,000 − limit price).
Read the last two columns together and the trade-off stops being abstract. A tight offset barely changes your loss but hands you a large chance of no fill. A wide offset buys near-certainty of filling — and by the time it is wide enough to survive a real move, the guaranteed loss has climbed to within touching distance of what a market order would have cost anyway.
That is the conclusion worth carrying: a stop-limit with an offset wide enough to always fill is a stop-market with extra steps and extra ways to fail. The figure below runs both ends of the table through the same price move.
Which price should trigger it — last, mark, or index?
This choice does not exist in stock trading and is skipped by almost every general order-type guide, but every major crypto venue asks you for it. The three references are defined in full on our mark price page; what matters here is what each one does to a stop-limit specifically.
| Trigger reference | What wakes your order | Failure mode for a stop-limit |
|---|---|---|
| Last price | The most recent trade on this venue | A one-venue wick triggers you at a price no other exchange saw — then price snaps back above your limit and the order never fills |
| Mark price | A smoothed multi-exchange value | Harder to trigger spuriously, but it lags — in a genuine fast drop it can wake the order after price is already through your limit |
| Index price | Spot across several exchanges | Furthest from what your own book is actually trading, so the gap between trigger and fillable price is widest |
Notice that the two references pull in opposite directions, and a stop-limit is punished by both. Last price triggers too eagerly; mark price triggers too late. A stop-market shrugs off either problem because it fills regardless. This is a structural reason stop-limits underperform as protective orders on perpetuals, independent of how you set the offset.
What breaks when it triggers but does not fill?
More than the obvious. The obvious part is that you still hold a losing position. The part people discover the expensive way is what happens to the other order attached to it.
On Bybit, the exchange's own spot documentation states that a paired take-profit / stop-loss order is cancelled as soon as the other one triggers — not when it fills. It warns directly that if price rebounds, the triggered limit order may never reach an executable level while the corresponding order has already been cancelled. Put those together and the failure state is a position with no stop and no target left on it, held by someone who believes both are still in place. (Bybit help documentation, as of August 2026 — behaviour differs by venue and changes over time; check your own exchange's current terms.)
Two more mechanics from the same documentation are worth knowing before you rely on this order type. Bybit's Entire Position exit mode closes with a market order and offers no limit-price field at all, so the stop-limit behaviour you configured on one screen may simply not apply on another. And its spot take-profit/stop-loss orders reserve your assets from the moment they are placed, not from the moment they trigger — which quietly reduces the balance you have available for anything else.
When is a stop-limit the wrong tool?
Whenever being out matters more than the price you get out at — which, for a protective stop, is nearly always. The reason you set a stop was to cap a loss you had already decided was your maximum. An order that may decline to execute is not capping anything; it is capping the loss only in the conditions where you did not really need the cap, and standing aside in the conditions where you did.
The specific cases where it is the wrong tool: thin books and small-cap pairs, where the gap between resting bids routinely exceeds any sane offset; weekends and low-liquidity hours; anything held through a scheduled event; and any position large enough that clearing it needs several price levels of depth. In all of these, take the slippage and accept the market fill.
Where it earns its place: entries rather than exits. Buying a breakout with a stop-limit is a reasonable trade — if price runs away past your limit, you miss the entry, and missing an entry costs you nothing but opportunity. That asymmetry is the whole point. Failing to fill on the way in is an inconvenience; failing to fill on the way out is a loss with no floor under it. It is also defensible for taking profit into strength, where the same logic applies: no fill simply means you are still in a winning trade.
And the condition under which this page's own advice is wrong: if you are trading a deep-book major, in normal hours, at a size the top of the book absorbs easily, and your genuine risk is a momentary wick rather than a sustained move, then a stop-limit with a sensible offset protects you from a bad print at almost no cost. The advice here is calibrated for the case that actually empties accounts — the sustained gap — not for the ordinary day.
FAQ
Is a stop-limit safer than a stop-loss market order? Safer on price, riskier on certainty. A stop-limit cannot fill below your limit price, but it can fail to fill at all. A stop-market always fills and accepts whatever the book gives. Which risk you prefer depends on whether an unfilled stop is acceptable to you — in a fast move, it usually is not.
What should I set the limit price to? Far enough below the stop that ordinary noise cannot skip it, close enough that the fill still matters. But note the trade-off: an offset wide enough to fill in every condition produces roughly the same fill a stop-market would have given, with extra ways to fail.
Why did my stop-limit trigger but never fill? Because price moved through your limit price before a counterparty met it. The trigger and the fill are two separate events. On some venues the paired take-profit order is cancelled the moment the stop triggers, so an unfilled stop-limit can leave a position with neither a stop nor a target attached.
Does a stop-limit protect me from a gap or a weekend move? No. Gaps are exactly the case a stop-limit does not cover. If price opens below your limit price and never trades back up to it, the order rests unfilled while the position keeps losing.