While exploring the answer to this question the question arose about whether or not the illusion of a creature type would have sufficed for magic requiring a given creature type to work.

Is an illusion of a dragon (or any other creature type) an acceptable stimulus to trigger a spell that specifies a dragon (or any other creature type), or does the actual creature of that type have to be present? As examples I'll use the spells Magic Mouth and Alarm, although there are other spells that require triggers.

All citations are from the SRD V_5.1

  1. I instruct the magic mouth to sing "See You In September" whenever an aberration gets within 30' of it. My clever gnome comrade sends the illusion of a gibbering mouther toward the magic mouth in order to play a prank on me. Does the illusion trigger the magic mouth, or do we need a real gibbering mouther to trigger that spell?

    Magic Mouth
    The triggering circumstance can be as general or as detailed as you like, though it must be based on visual or audible conditions that occur within 30 feet of the object. For example, you could instruct the mouth to speak when any creature moves within 30 feet of the object or when a silver bell rings within 30 feet of it. (p. 161)

  2. I set the trigger for an alarm to alert me if a dragon enters the alert area. An illusionist who has been our party's nemesis moves an illusion of a wyvern (using the spell major image) into that area while doing reconnaissance before sending his minions after us.


    Until the spell ends, an alarm alerts you whenever a Tiny or larger creature touches or enters the warded area. When you cast the spell, you can designate creatures that won’t set off the alarm. You also choose whether the alarm is mental or audible. (p. 114)

Bottom Line

Does the illusion suffice as criteria for either of the spells to trigger, or does a spell with such a trigger need the real thing?

Put in other words, can spells be fooled by illusions? (Thank you @Mindwin).


  1. Creature Types are (p. 254, 255): Aberrations, Beasts, Celestials, Constructs, Dragons, Elementals, Fey, Fiends, Giants, Humanoids, Monstrosities, Oozes, Plants, Undead.
  2. Gibbering Mouther (p. 314): Medium aberration, neutral
  3. Wyvern (p. 356): Large dragon, unaligned
  • 2
    \$\begingroup\$ As a side note, identifying aberrations might be beyond the bounds of "visual conditions". The spell probably can't tell that a neogi is an aberration but a drider isn't, for example. \$\endgroup\$
    – Miniman
    Commented Jan 11, 2018 at 22:51
  • \$\begingroup\$ @Miniman That creature type may be a bad example. Might I better use beast for that example? The point is to identify type as the issue, and whether real or fake is discernable ... the trigger conditions seem to have a lot of latitude ... \$\endgroup\$ Commented Jan 11, 2018 at 23:06
  • 3
    \$\begingroup\$ @KorvinStarmast Honestly, I'm not sure. I mean, the visual distinction between, say, a dire wolf and a worg seems really miniscule. But, thinking about it, I don't think I'd let a character tell what a creature's type was just by looking, either. \$\endgroup\$
    – Miniman
    Commented Jan 11, 2018 at 23:15
  • 2
    \$\begingroup\$ @KorvinStarmast Yeah, it's a weird problem. \$\endgroup\$
    – Miniman
    Commented Jan 12, 2018 at 2:10
  • 2
    \$\begingroup\$ I always laugh a little when a conversation like "How would a magic spell be able to discern something about a fictional, magical creature in a fantasy universe" is restricted to visual clues. \$\endgroup\$
    – Luke
    Commented Jan 12, 2018 at 4:25

2 Answers 2


Magic Mouth should trigger, but not Alarm

As always: many things are ultimately up to the DM, but the term "visual or audible conditions" in Magic Mouth implies only a visual/audible 'sensor'. Illusions should be able to fool visual/audible sensors, while a silent/invisible aberration should not trigger the Magic Mouth (by RAW it did not meet the visual/audible conditions). Giving the Magic Mouth spell greater detection abilities than this would seem unwarranted, and could lead to unexpected uses or abuses.

Conversely, the sensor for the Alarm spell somehow detects when a "Tiny or larger creature touches or enters the warded area" without mentioning sight/hearing. I'd use strict RAW here: an illusion wouldn't fulfill the triggering condition, but an invisible creature would.

For other spells - I'd similarly base the trigger detection abilities off their description. Examples:

  • Detect Magic: "see a faint aura around any visible creature or object" implies that you wouldn't see it around an illusion (per Jeremy Crawford), but the trigger detection should still be fooled by Invisibility (an illusion spell) due to the "visible" clause in the above statement.
  • Glyph of Warding can activate "under certain circumstances or according to physical characteristics (such as height or weight), creature kind (for example, the ward could be set to affect aberrations or drow), or alignment". So we know that it can detect all those things. But having it detect say, 'someone that lied in the last 5 minutes' (to serve as a lie detection spell) wouldn't seem warranted.
  • 2
    \$\begingroup\$ Ah, it appears that a general answer might require that one investigate in detail a given spell due to specific > general. I am glad that I chose those two spells, given your and Rubik's replies. Luck, not skill, in that choice. \$\endgroup\$ Commented Jan 11, 2018 at 20:31

Magic mouth could trigger, but alarm will not

General rule: illusions are not creatures and thus will not trigger if the spell requires a creature regardless of creature type

Illusions are visual spell effects that might look, sound, or even act like creatures. But, they are in fact still just magic effects masquerading as creatures.

Jeremy Crawford has confirmed this in a question about detect magic:

Detect magic lets you see an aura only around a visible creature or object, not around an illusion.

And also in this question about Invoke Duplicity:

The illusion of Invoke Duplicity isn't a creature, an enemy, or an ally. It doesn't interact with Sneak Attack.

Note that in these two rulings, Jeremy has said that illusions cannot be affected by abilities or spells that specifically target creatures. Creature is a game term with mechanical implications. And illusions explicitly do not qualify as creatures.

If a spell targets/affects only specifically creatures then you may not use an illusions to qualify.

It also doesn't matter if the spell targets a creature or a creature subtype, it still doesn't work with illusions in general.

Application to specific spells

Moving from the general rule, I'll apply it to the two specific spells you asked about.

  1. Magic Mouth might get fooled depending on the way the trigger is worded

    First, note that magic mouth does not call out creatures anywhere in its spell it will be dependent entirely on how the caster of the spell words their triggers. Since magic mouth depends on auditory or visual triggers it can be fooled depending on how the trigger is worded.

    Generally, as long as magic mouth sees or hears something that resembles the creature that you specify in the trigger, it will indeed trigger. As an example: a trigger worded as "anything that looks like a zombie" will definitely trigger off a zombie illusion.

  2. Alarm cannot be fooled by an illusion

    However the mechanics of alarm works, it can detect creatures and creatures only. Alarm does not depend on sight and thus is not fooled by visual illusions. Instead, it does exactly what the description says it does: triggers if a creature enters the area.

  • 1
    \$\begingroup\$ Glad to see the Crawford rulings folded in, as well as the specific spell cases. \$\endgroup\$ Commented Jan 11, 2018 at 21:41
  • 1
    \$\begingroup\$ @KorvinStarmast Since you asked general questions in your OP I thought I'd also give you the general ruling for completeness. \$\endgroup\$ Commented Jan 11, 2018 at 22:01
  • 1
    \$\begingroup\$ Sometimes, when asking a question, one may not be aware that it may rely on specifics. That's part of the process, and is a Good Thing(TM). \$\endgroup\$ Commented Jan 11, 2018 at 23:05

You must log in to answer this question.

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