Every guide to writing affirmations gives you the same rule, usually as the first one. Should affirmations be in present tense? Yes, they say, always. Not one of them cites anything, and having gone looking, I can see why.
Should affirmations be in present tense: the honest answer
Nobody has tested it. There is a large literature on self-talk, and within it two specific grammatical questions have been put to experiment. Tense is not one of them. What makes this worth a page rather than a shrug is that the two questions that were tested have clear results, and neither is the one you have been told about.
Where the rule came from
The instruction usually arrives as part of a set: personal, positive, present tense. It reads like something distilled from research. In truth it is distilled from other people's blog posts.
The lineage runs back through a century of self-help writing to the New Thought movement, where the reasoning was explicitly about belief rather than measurement. Each generation restated it, and by now the rule is old enough that nobody thinks to ask where it started.
The justification you will meet most often is that your subconscious hears I will as a confession that you are not that yet, and quietly confirms the gap. A tidy story, and one that describes a mechanism nobody has measured. That is the same move we take apart elsewhere on this site.
One twist is worth flagging. A claim circulates that future-tense lines work better for people with low self-esteem, and it is usually traced to a well-known 2009 study. That study's own summary describes different conditions, so the claim has drifted somewhere between the paper and the page you read it on. We are not going to repeat it as fact.
The grammar question that was tested, and fell over
In 2010, researchers published something genuinely surprising in Psychological Science. Asking yourself a question beat telling yourself a statement.
People who wrote Will I before a task solved more anagrams than people who wrote I will. It worked even when the writing was disguised as an unrelated handwriting exercise. Four experiments, a neat mechanism, and an irresistible headline.
Then other labs tried it. Direct replications found nothing, and a meta-analysis pooling six replication studies found no significant difference between asking and telling.
We include this precisely because it is inconvenient. It is the one grammatical claim in this area that was properly tested and widely repeated, and it did not survive. If you have seen the Will I trick recommended, that is where it came from and that is what happened to it.
The grammar question that held up, modestly
The other tested question is person. Not when the sentence happens, but who it is addressed to.
In 2014, seven studies with 585 participants compared people who worked through a stressful task using I with people who used their own name or you. The second group performed better in front of independent raters who did not know which group anyone was in, reported less distress, and spent less time going over it afterwards.
The explanation offered is distance. Your own name puts a half step between you and the situation, roughly the distance you already have when a friend describes their problem to you and the answer seems obvious.
Now the part that most pages quoting this leave out. A 2023 review pooled 25 experiments and 2,397 participants and found the effect real but small, with a confidence interval running from 0.05 to 0.33. The lower end of that range is close enough to zero to be worth saying out loud. This is a modest, reliable effect, not a trick.
Three versions of the same line
Take one sentence and change one grammatical thing at a time. The first column is what the rule demands. The second is the change with evidence behind it.
I am calm before difficult conversations.
Present tense, first person. The approved form, and the easiest to argue with.
Sam, you have done this before and it went fine.
Same content, addressed to you. The one change that has been tested.
I am confident in interviews.
A claim your memory can disprove in under a second.
You know the first thing you are going to say.
Names a fact rather than a trait, and keeps the distance.
I will be someone who exercises.
Future tense. Forbidden by the rule, and perfectly usable.
You are going out at six, the same as Tuesday.
Tense barely matters here. The specifics are doing the work.
What to actually do with this
Stop optimising the tense. There is nothing there to optimise, and the effort would be better spent on the claim the sentence makes.
If you want to change one grammatical thing, change the person. Use your own name or you. It costs nothing, it is the only change in this area with repeated support, and the worst case is that it feels odd for a week.
Beyond that, use whichever tense you can say without your mind producing a rebuttal. For something already true, the present works. For something you are moving towards, the future is honest and the present is a small lie you will notice every time.
One thing the 2009 study does support, and it points away from the rule rather than towards it. The harm it found for people with low self-esteem came from repeating a strongly positive present-tense declaration. Exactly the form the rule insists on, given to exactly the readers most guides are written for.
When time is tight
- Step one
Tense is untested.
The oldest rule in the category, and the only one nobody has put to an experiment.
- Step two
Asking instead of telling failed.
A famous 2010 result that six replication studies could not reproduce.
- Step three
Person holds up, modestly.
Use your name or you. Real effect, small, repeatedly found.
The limits of this
Switching from I to your own name is free and has more behind it than any other wording advice in this category, and that is worth knowing given how confidently the alternatives are sold. The effect is also small, measured mostly in short laboratory tasks like public speaking, and it won't carry a sentence whose content you don't believe. Nothing on this page says tense cannot matter. It says nobody has checked, and that a rule repeated for a century without a test should not be treated as settled. If a line keeps leaving you worse than before you said it, the grammar is not the problem and no rearrangement of it will help.
Sources and further reading
These references support the research and practical guidance discussed in this article.
- Self-talk as a regulatory mechanism: how you do it mattersKross and colleagues, Journal of Personality and Social Psychology, 2014. Seven studies, 585 participants. Referring to yourself by name or as you, rather than as I, produced better speech performance judged by independent raters, less distress, and less going over the event afterwards.
- The effectiveness of self-distanced versus self-immersed reflections among adults: systematic review and meta-analysisMurdoch and colleagues, Stress and Health, 2023. Twenty-five experiments, 2,397 participants. Pooled effect g = 0.19, confidence interval 0.05 to 0.33. Real, small, and with a lower bound close to zero.
- Motivating goal-directed behavior through introspective self-talk: the role of the interrogative form of simple future tenseSenay, Albarracín and Noguchi, Psychological Science, 2010. Writing Will I rather than I will improved anagram performance and intentions to exercise. Widely cited, and see the next entry.
- I will use declarative self-talk, or will I? Replication, extension, and meta-analysesVan Raalte and Cornelius. Direct replications plus meta-analysis of six replication studies found no significant difference between the interrogative and declarative forms. The reason this page treats the 2010 result as unsupported.
- Positive self-statements: power for some, peril for othersWood, Perunovic and Lee, Psychological Science, 2009. The harm found for people with low self-esteem came from repeating a present-tense declaration, which is the form the rule insists on.
