According to the 3.5 PHB (p. 309), a helpless (e.g., sleeping) target is treated as having Dex 0, and has a "-5 modifier". On p. 153, it further explains that a helpless defender can't use any Dexterity bonus "to AC. In fact, his Dexterity score is treated as if it were 0 and his Dexterity modifier to AC as if it were -5." (Emphasis added.) It also states that a coup de grace automatically hits and does critical damage.

What I can't seem to find is any discussion of whether a sleeping or otherwise helpless target gets a Reflex saving throw. I'm particularly interested in whether a helpless target gets a Reflex save vs. dragon breath weapons.

It seems to me that a helpless, particularly sleeping, target won't even know it's under attack, and thus won't be able to "escape by moving quickly" when dragon breath, a spell attack or the like comes at it.

How do the rules deal with this situation? If they don't address it, how can it best be resolved?


5 Answers 5


This is not strictly RAW, but it is from an article called Rules of the Game: Actions (Part Two) by Skip Williams on the WotC website.

It's usually best to allow... opposed checks to resist being tripped or grappled, unless they're helpless (any ability score reduced to 0, paralyzed, petrified, or unconscious). It's worth noting, however, that even a helpless character can make a Reflex save (albeit with a -5 penalty for an effective Dexterity score of 0). That's because the game generally favors at least some involvement from the player -- and a chance for a miraculous escape -- whenever a character is in danger.

So Jonathan Drain's comment is dead on. The saving throw doesn't map directly to dodging out of the way, it's a metagame mechanic. A good way to think of it is that the "reflex" part of the save is best used after the roll to narrate how the character avoided death.

  • \$\begingroup\$ Addendum to my answer: personally I'm torn between SevenSidedDie's approach and the approach that allows a chance for miraculous escape. As a DM and impartial arbiter of the game world, I prefer the idea that some sort of baseline "reality" should be consulted before rolling dice. But as a player, I prefer the idea that there should be some chance for survival, no matter what the odds. Split the difference and give penalties to the roll, is my advice. :) \$\endgroup\$
    – cr0m
    May 21, 2012 at 19:16
  • 2
    \$\begingroup\$ Despite having an answer here that differs, I'm giving this a +1 anyway! 3e can be understood as a hybrid game of old and new sensibilities, and there's definitely room for the old style of saves where they are "by rights you're gonna die, but here's a chance for a miraculous escape." I think giving saves to helpless characters as such a Hail Mary roll is a totally legitimate choice as part of a particular DM's style. \$\endgroup\$ May 21, 2012 at 20:53
  • \$\begingroup\$ Thanks! FWIW I upvoted yours too, because "reality" based play is very, very satisfying (and avoids a lot of the rules exploits/lawyering endemic to modern D&D). \$\endgroup\$
    – cr0m
    May 22, 2012 at 5:43
  • \$\begingroup\$ I think this is the way to go; at a -5 dex modifier a successful save is really representative of miraculous luck. A character's prowess can certainly tip the scales, but at the end of the day, all saves are a measure of how lucky they were. \$\endgroup\$
    – Llepwryd
    Dec 20, 2013 at 19:26

D&D3.x did try to standardise and clarify a lot of things, but it was still built on a foundation that assumed humans would be applying the rules with common sense, and with sensitivity to the needs of the fictional situations that could not be predicted by the rules.

You could just give them the Reflex save, and you'd be following the literal rules. That's one option.

Your instinct that common sense says otherwise is good though. They can't move out of the way, so giving them the Reflex save doesn't agree with common sense. Just denying it when it seems to not apply is a fine option that agrees with the spirit of the rules.

However, Jonathan Drain is right: a Reflex save is usually used to stand in for more than just physical dexterity. It often serves as a stand-in for luck, for circumstantial advantages of terrain not directly modelled by explicit declarations, for the variations in an effect that covers an area, and so on. Ignoring this unspoken meaning of Reflex saves for the sake of common sense wouldn't be entirely accord with good sense either.

The best option then, is to make a judgement call and decide what does need to be settled by the dice, and roll that. Only you will be able to judge the exact details of the precise situations that come up in your game, but some simplified examples will round out this answer:

  • Sleeping in a tight circle around a campfire, and hit by a fireball? They're all in trouble, no save.
  • Caught in the AoE of a blast of dragon fire directed at someone else while sleeping? Maybe they get lucky: roll a d20, high for half damage, low for full damage.

Trust your instincts as a DM: that's why you're there instead of a computer.

  • 5
    \$\begingroup\$ I can just see it now- Player is hit by sleep spell, falls on a pressure plate that activates a needle trap. Player immediately tosses and turns in sleep to more comfortable position with the unintentional added bonus of barely avoiding a poison dart to the spleen. \$\endgroup\$
    – Holman716
    Aug 16, 2011 at 14:04
  • \$\begingroup\$ That's the spirit of creative narration! Now try harder. \$\endgroup\$
    – cr0m
    Dec 20, 2013 at 5:52

Yes they do get a reflex save but they do not get to use the Dex bonus (if any, penalties would still apply tho I suspect). A sleeping person is officially helpless for actual effects that apply, etc.

This thread discusses the issue somewhat.

  • 1
    \$\begingroup\$ Thanks for the reply and link to the related thread. I'm trying to reconcile the RAW, as you refer to it, with common sense (probably a dangerous exercise in connection with D&D). It seems to defy common sense to be able to evade a huge cone of dragon breath while you're asleep. Could it be that having a high Reflex save actually means that your "reflexes" are so developed that you actually can pick up on incoming dangers while asleep (even magically asleep) and then get out of the way? \$\endgroup\$ Aug 14, 2011 at 5:55
  • 3
    \$\begingroup\$ It's still possible to avoid injury even if you can't move. You might get lucky and the brunt of the blast happens to miss you, or your armour might protect you, or the dragon might have been intentionally directing his blast at someone else, or the enemy wizard may have cast his spell poorly and it didn't affect you as badly as it might, or you're such a great adventurer that you instinctively fell asleep in a defensible part of the room. \$\endgroup\$ Aug 14, 2011 at 6:46
  • \$\begingroup\$ That makes sense. As a DM, I'd have to say that denying Reflex saves to sleeping PC's, for instance, would be very unbalancing. Likewise with NPC's and monsters. \$\endgroup\$ Aug 14, 2011 at 7:28
  • \$\begingroup\$ It's highly likely the part of the logic of allowing reflex saves is that a sleeping character can still make Fortitude/Will saves and have a situation where you can't make one of three is fairly rare (if it happens at all?) in 3.x Easier to just give a penalty to most affected save. \$\endgroup\$
    – mirv120
    Aug 14, 2011 at 11:15
  • \$\begingroup\$ I disagree. If your dex is (effectively) 0, you CANNOT MOVE. I hate to say it, but if you are in the line of fire and cannot move, you are getting hit by whatever it is. If you cannot roll over because someone with a sword wants to coup de grace you, you should not be able to roll over to dodge a dragon's breath. \$\endgroup\$
    – Pulsehead
    Aug 14, 2011 at 12:59

Helpless targets do roll a Reflex save.

Saves are much like AC: numerical stats that measure your ability to avoid being affected by the environment. The main difference between saves and AC is that the latter is a static number whilst saves must be rolled, but mechanically they both affect the outcome of a d20 roll.

A helpless target is entitled to use his AC (in order to circumvent this you have to take a coup de grace action, which is riskier and takes more time than a regular attack). So, a helpless character in the midst of a dragon's breath is entitled to use his Reflex bonus for half-damage (albeit applying a -5 Dex modifier instead of his usual one).

If you want to introduce a house rule, I'd suggest to extend the coup de grace action to those attacks that grant a save too. So, for example, a dragon may choose to coup de grace the helpless character with its breath weapon; doing so will take a full-round action, will provoke attacks of opportunity, will require the dragon to be adjacent to the target, but will result in an automatic failure of the required Reflex save.


The 3/.5 player's handbook, when talking about saves, specifically mentions you CANNOT take a reflexive save if you're flatfooted (reflex is the only save with this restriction), unless you have a class ability/feat that says otherwise. Sleeping or otherwise helpless creatures are flatfooted.

  • 5
    \$\begingroup\$ Could you provide a page reference and quote here? The only section on saves that I can find in the 3.5 PHB is on page 136, and it doesn't mention anything there about not getting reflex saves while flat-footed. \$\endgroup\$
    – DuckTapeAl
    May 18, 2012 at 5:09
  • \$\begingroup\$ giantitp.com/forums/… \$\endgroup\$ Apr 22, 2015 at 18:19
  • \$\begingroup\$ Were this true, wizards and other casters would be even more powerful than they already are … except that spells that deal damage and mandate Ref saves (e.g. fireball) normally suck. So this would encourage blasting wizards to A) ambush enemies and B) pump their initiatives (hint: they already were doing both) so as to maximize the impact of their spells. I don't know if this would be a good change or a bad one, but it's definitely not the rules. (Perhaps this is being confused with immediate actions or attacks of opportunity, both usually unavailable when flat-footed.) \$\endgroup\$ Sep 7, 2020 at 11:37

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .