Stata Error r(111): '[variable] not found' — Complete Fix Guide

Your variable exists. You created it two lines ago. You can see it in the Variable window. You run the command and Stata says:

variable eduaction not found
r(111);

Error r(111) means Stata cannot find a variable you referenced. It does not exist in the current dataset — or at least, not under the name you typed. This guide covers every documented reason this happens and shows the fix for each.

All examples tested in Stata 18 SE. Compatible with Stata 15+.

Quick Answer

Error r(111) means the variable name you typed does not match any variable in memory. The most common causes are:

1. Typo in the variable name
2. Case-sensitivity mismatch (Income vs income)
3. Variable was dropped or never created
4. Variable is in a different data frame (Stata 16+)
5. Forgot i. prefix for factor variables
6. Time-series operator without tsset

The fastest diagnostic: run describe to see every variable in your dataset, then lookfor to search by keyword.

1* See all variables
2describe
3
4* Search for a variable by keyword
5lookfor wage
6lookfor educ

Cause 1: Typo in the Variable Name

The most common cause, period. One wrong letter and Stata cannot find the variable. Common patterns: transposed letters (edcuation), missing letters (educaton), or extra characters (education1 when the variable is education).

regress wage eduaction experience
* "eduaction" instead of "education"

regress wage education experience

The fix

Run lookfor with part of the variable name. It searches both variable names and labels:

1lookfor educ

              storage   display    value
variable name   type    format    label      variable label
─────────────────────────────────────────────────────────────
education       float   %9.0g                Years of education

Cause 2: Case Sensitivity

Stata variable names are case-sensitive. Income, income, and INCOME are three entirely different variables. If your dataset hasIncome and you type income, you get r(111).

summarize income
* But the variable is "Income" (capital I)

summarize Income
* Match the exact case from describe

The fix

Run describe and check the exact casing. If you want to permanently fix inconsistent casing, rename the variable:

1* See exact variable names
2describe, simple
3
4* Rename to lowercase (common convention)
5rename Income income
6rename FIPS fips

Cause 3: Variable Was Dropped Earlier

If your do-file runs drop or keep before the line that references the variable, the variable no longer exists. This is especially common in long do-files where you clean data in one section and analyze in another.

1* Data cleaning section
2use "raw_data.dta", clear
3drop if missing(wage)
4drop race ethnicity religion    "stata-comment">// dropped some variables
5
6* ... 200 lines later ...
7
8* Analysis section — oops, "race" was dropped
9regress wage education i.race   "stata-comment">// r(111): race not found

The fix

Search your do-file for drop and keep commands. If you need the variable later, either don’t drop it, or reload the data. You can also use preserve/restoreto make destructive changes safely:

1preserve
2    keep wage education experience
3    regress wage education experience, robust
4restore
5* All variables are back

Cause 4: Variable Is in a Different Frame

Stata 16+ supports multiple data frames in memory. If you loaded data into a named frame but your command runs against the default frame, Stata won’t find the variable.

1* Load data into a named frame
2frame create analysis
3frame analysis: use "mydata.dta"
4
5* This fails — default frame has no data
6summarize wage                    "stata-comment">// r(111)
7
8* This works — specify the frame
9frame analysis: summarize wage
10
11* Or switch to the frame
12frame change analysis
13summarize wage                    "stata-comment">// now it works

Cause 5: Abbreviation Ambiguity

Stata allows variable name abbreviation — you can type edu instead of education if no other variable starts with edu. But if multiple variables share the same prefix, the abbreviation is ambiguous and Stata throws r(111).

1* Two variables starting with "edu"
2describe edu*

              storage   display    value
variable name   type    format    label      variable label
─────────────────────────────────────────────────────────────
education       float   %9.0g                Years of education
edu_level       byte    %9.0g     edlbl      Education level

1* Ambiguous abbreviation — Stata can't tell which you mean
2summarize edu         "stata-comment">// r(111)
3
4* Be specific
5summarize education   "stata-comment">// works
6summarize edu_level   "stata-comment">// works

Cause 6: String vs Numeric Confusion

If you try to use a numeric operator on a string variable (or vice versa), the error can sometimes manifest as r(111) rather than a type mismatch error — especially inif conditions with incorrect comparison syntax.

* If state_id is a string variable
summarize wage if state_id == 6

* Use quotes for string comparison
summarize wage if state_id == "6"
* Or if you need numeric: destring first
destring state_id, replace
summarize wage if state_id == 6

Cause 7: Time-Series Operators Without tsset

Time-series operators like L. (lag), F. (lead), and D. (difference) require that your data is declared as time series with tsset or xtset. Without this declaration, Stata interpretsL.gdp as a variable literally named L.gdp — which doesn’t exist.

1* Forgot to tsset
2regress gdp_growth L.gdp investment    "stata-comment">// r(111): L.gdp not found
3
4* Fix: declare time structure first
5tsset country_id year
6regress gdp_growth L.gdp investment    "stata-comment">// now works

variable L.gdp not found
r(111);

Cause 8: Factor Variables Without i. Prefix

If you want to include a categorical variable as a set of dummies in a regression, you need the i. prefix. Without it, Stata treats the variable as continuous. But certain contexts — like interactions — require the prefix, and omitting it can produce r(111) in some command contexts.

1* If "industry" is a numeric categorical variable (1, 2, 3, ...)
2* This may give unexpected results or errors
3regress wage education industry
4
5* This creates dummies correctly
6regress wage education i.industry
7
8* Interactions also need the prefix
9regress wage education i.industry##c.experience

Cause 9: Referencing _n and _N Incorrectly

_n (current observation number) and _N (total observations) are system variables. They cannot be used in all contexts — for example, you cannot use them in by prefix contexts without understanding that they reset per group.

1* _n and _N are not regular variables
2describe _n         "stata-comment">// r(111) — _n doesn't show up in describe
3
4* But they work in expressions
5gen obs_number = _n
6gen total_obs = _N
7
8* With by-groups, _n and _N are within-group
9bysort state: gen state_obs = _n
10bysort state: gen state_total = _N

Cause 10: After clear or use

Running clear removes all data and variables from memory. Runninguse replaces the current dataset. If you reference a variable from the previous dataset after either command, you get r(111).

1use "wages.dta", clear
2gen log_wage = ln(wage)
3
4* Now load a different dataset
5use "demographics.dta", clear
6
7* This variable no longer exists
8summarize log_wage    "stata-comment">// r(111): log_wage was in the wages data

The fix

If you need variables from multiple datasets, either merge the datasets first, or use frame (Stata 16+) to keep both datasets in memory simultaneously.

Cause 11: preserve/restore Scope

Variables created between preserve and restore are discarded when restore executes. If you created a variable inside a preserve block and try to use it after restore, it’s gone.

1preserve
2    gen temp_indicator = (wage > 50000)
3    tab temp_indicator
4restore
5
6* temp_indicator no longer exists
7summarize temp_indicator    "stata-comment">// r(111)

1preserve
2    keep if wage > 50000
3    count
4    local high_wage_n = r(N)
5restore
6
7display "High-wage observations: `high_wage_n'"

Debugging Checklist

When you hit r(111), work through this list:

1. Run describe. See every variable in the current dataset.
2. Run lookfor keyword. Search variable names and labels for what you’re looking for.
3. Check the spelling. One wrong letter is enough.
4. Check the case. Wage and wage are different.
5. Check for drop/keep. Did you remove the variable earlier in your do-file?
6. Check tsset/xtset. Time-series operators need a declared time structure.
7. Check the active frame. Run frame to see which dataset is active.

FAQ

What does r(111) mean in Stata?

Error r(111) means Stata cannot find a variable you referenced. The variable does not exist in the current dataset — either it was never created, was dropped, is misspelled, or lives in a different frame.

Why does Stata say variable not found when I can see it?

The most common reason is a case mismatch. Stata is case-sensitive: Income andincome are different variables. Run describe to see the exact names. Another possibility: the variable is in a different frame (Stata 16+).

How do I check which variables exist in my dataset?

Run describe to list all variables with their types and labels. Run ds for a compact list of names only. Run lookfor keyword to search both names and labels.

Can r(111) happen after a merge?

Yes. If you used keepusing() during merge, only the specified variables from the using dataset are brought in. If you expected a variable that wasn’t in keepusing(), it won’t be in the merged result and you’ll get r(111).