Arcade String Functions: What They Do and How to Use Them

Formulas & functions · Updated November 22, 2026

In Esri Arcade, string functions work with text: they can join values, search for characters, extract part of a string, change letter case, or split text into pieces. Use functions such as Concatenate, Find, Mid, Replace, and Split in an Arcade expression when you need to clean, inspect, or present text from a feature or another data source.

What “string functions” means in Arcade

Arcade is Esri’s expression language for working with geographic data and building behavior in ArcGIS products. It is used in places such as pop-ups, labels, forms, and attribute rules, with the available data and capabilities depending on the expression profile. A string is simply a sequence of text characters, such as a street name, a category, or a note stored in an attribute.

“Arcade” can also mean a video-game venue, but “string functions Arcade” usually refers to text-processing functions in Esri Arcade. These are not automatically the same as JavaScript string methods. Arcade has syntax that may look familiar to JavaScript users, but an expression should use Arcade’s documented functions rather than assuming methods such as .toUpperCase() or .substring() are available.

Functions are useful when the stored value is not already in the form a map, form, or calculation needs. For example, an expression can combine a town and state for a label, standardize a category’s capitalization, or extract a prefix from an asset code. The result is still governed by the profile in which the expression runs: a pop-up can return a different kind of result than a label expression can display.

Common Arcade string functions

These functions cover many routine text tasks. Function names and arguments matter, so check the order shown in each example:

  • Concatenate(values, separator) joins values into one string. For example, Concatenate([$feature.City, $feature.State], ", ") combines two feature attributes with a comma and space.
  • Find(searchText, text, startPos) returns the position where a search string first occurs, or -1 if it is not present. The optional starting position lets a search begin later in the text.
  • Left(value, charCount) returns characters from the beginning of a string; Right(value, charCount) returns them from the end.
  • Mid(value, startIndex, charCount) extracts a specified number of characters beginning at a position in the string.
  • Lower(value) and Upper(value) convert letters to lowercase or uppercase. Proper(value) applies title-style capitalization.
  • Replace(value, searchString, replacementString) substitutes matching text. Trim(value) removes whitespace at the beginning and end of a string.
  • Split(value, delimiter) divides a string at a delimiter and returns an array of pieces.
  • StartsWith(value, searchString) and EndsWith(value, searchString) check whether text begins or ends with a particular string.
  • Text(value) converts a value, such as a number or date, to text. Optional formatting can be used where appropriate.

Arcade also supports joining strings with the + operator in suitable expressions. Concatenate is useful when joining an array of values or when a separator should be inserted consistently. For a readable display string, a separator such as " - " or ", " often makes the result easier to scan than joining words without punctuation.

Extracting and locating text

Left, Right, and Mid are handy when a value follows a predictable pattern. For example, imagine an inventory field containing BLDG-0472. If the prefix always uses five characters, Left($feature.AssetCode, 5) returns BLDG-. If the number of characters to keep can vary, a fixed-position extraction may be brittle; finding the hyphen first can be a better approach.

Find reports a zero-based position: the first character is at position 0. For example, in Find("-", "BLDG-0472"), the hyphen is at position 4. A result of -1 means the search string was not found, so an expression should account for that possibility before using the result as an extraction position. Otherwise, a value that does not match the expected pattern may produce an unintended result.

Mid also uses a starting index, so a small test with representative values is worthwhile before applying it across a layer. If an identifier has a fixed prefix and fixed-length suffix, a fixed index may be simple and reliable. If records contain different prefixes, separators, or missing parts, combine searching with checks for the expected separator instead of assuming every record has the same layout.

Replacing, splitting, and assembling strings

Use Replace for a known text substitution, such as changing a legacy abbreviation in a display value. Its arguments are the original value, the text to look for, and the replacement text. It is a text operation, not a general data-cleaning guarantee: decide whether the substitution should apply to every occurrence and whether capitalization or punctuation variants need separate handling.

Split is useful when a field stores several parts separated by a consistent character. For instance, Split("North|Central|South", "|") returns an array with three entries. The pieces can then be referenced by their array positions, with the first entry at index 0. Before relying on a position, check that the source actually contains enough pieces; a missing delimiter or an extra delimiter can change the structure.

For the reverse task—creating a display string from separate attributes—use Concatenate. An expression such as Concatenate([$feature.Street, $feature.City], ", ") produces a readable street-and-city value when those attributes contain usable text. If the source attributes can be null or empty, decide how the output should look in those cases. You might omit a missing component, substitute a clear placeholder, or keep a separator only when both values exist. That decision is a data-display rule, not something a separator can resolve automatically.

Case, whitespace, and matching

Inconsistent letter case can make categories look like different values to a reader even when they mean the same thing. Lower and Upper can standardize text for comparisons or display. For example, applying Lower to a category before a comparison can help avoid treating Road and ROAD as separate spellings. Use the same normalization approach on both sides of a comparison.

Proper can make a name look more readable, but automatic title-style capitalization is not a substitute for authoritative name data. It may not preserve the intended capitalization of abbreviations, brands, or names such as McDonald. When exact spelling matters, correct the underlying attribute or maintain a curated display value rather than transforming it blindly.

Trim addresses leading and trailing whitespace, a common cause of confusing comparisons. A value that appears to be West may contain an unnoticed space after the final letter. Trimming can remove that outside whitespace, but it does not remove spaces between words or repair every kind of inconsistent punctuation. For matching, be explicit about which differences should count as equivalent: case, extra spaces, hyphens, and accents are separate issues.

Converting values to text

Attributes do not all begin as strings. A numeric field might hold a count, and a date field holds a date value. Text converts a value to text when a display expression needs to combine it with words or other text. For example, a label can join a fixed phrase with the text representation of a feature’s numeric field. Conversion for display does not change the stored field’s data type.

Formatting matters. A number intended as a whole-number count should not be presented with unnecessary decimal places, while a date may need a specific date format for its audience. Arcade’s Text function supports formatting options, but the correct format depends on the value type and the intended output. Check the Arcade documentation for the supported format pattern for that type and test the result in the target profile; date and number formatting are not interchangeable.

It is also useful to distinguish a missing value from a literal word such as "Null". An absent attribute is not necessarily the same as an empty string, and converting or concatenating values without considering missing data can create labels that look incomplete. Arcade provides functions such as IsEmpty for checking values. Build the intended missing-value behavior into the expression rather than assuming every feature has a complete record.

Choosing the right function and testing an expression

Start with the shape of the data and the result you need. Use Concatenate to assemble a label, Replace for a targeted substitution, Split for delimiter-separated parts, and Find when a separator’s position is not fixed. Use Left, Right, or Mid when the character positions or lengths are predictable. For simple case changes or outside whitespace, use Lower, Upper, or Trim.

Before publishing an expression, try it against ordinary records and edge cases. Include examples with a missing field, an empty string, a separator at the start or end, and text that does not contain the expected search term. Confirm that the result has the type the expression profile expects. Arcade is used across ArcGIS apps, but profiles do not all expose identical globals, data access, or return requirements; an expression that works in one context may need adaptation in another.

When an expression is used for labels or pop-ups, consider whether transformed text should be presentation-only. Changing capitalization in an expression can make a map look consistent without correcting the underlying attribute. If other workflows depend on clean, standardized values, address the data itself through an appropriate editing or data-management process rather than relying solely on a display expression.

Common mistakes and misconceptions

  • Using a JavaScript method by habit: Arcade resembles JavaScript in places, but its available API is specific. Use Arcade string functions documented for the language and profile.
  • Assuming positions start at one: String and array positions used by functions such as Find, Mid, and array indexing are zero-based.
  • Ignoring a failed search: Find returns -1 when the search text is absent. Check for that outcome before using a position to extract text.
  • Expecting automatic cleanup: Trim removes outside whitespace; it does not standardize every internal space, punctuation mark, or spelling difference.
  • Assuming Arcade changes the source data: A string function changes the expression’s result. It does not, by itself, edit the attribute stored in the feature.
  • Forgetting profile differences: The functions and data accessible in a labeling expression, pop-up, form, or attribute rule can differ. Confirm the target profile before reusing an expression.

Frequently Asked Questions

What is the Arcade function for joining strings?
Concatenate joins values into one string and can insert a separator. For example, it can combine a city and state with a comma and space. The + operator can also join text in suitable expressions.

How do I get part of a string in Arcade?
Use Left for characters at the beginning, Right for characters at the end, or Mid for a section starting at a chosen index. These functions work best when the source layout is predictable.

What does Find return when text is not present?
Find returns -1 when it cannot locate the search text. Its successful result is a zero-based position, so the first character is position 0.

Does Trim remove spaces between words?
No. Trim removes whitespace at the beginning and end of a string. It does not collapse repeated spaces between words or fix other formatting inconsistencies.

Are Arcade string functions the same as JavaScript string methods?
No. Arcade has its own documented functions and rules. Do not assume that a JavaScript method such as .toUpperCase() or .substring() will work in an Arcade expression.

Do Arcade expressions permanently change attribute values?
Usually, a string function only produces a result for the expression’s context, such as a label or pop-up. It does not edit the source attribute unless the expression is being used in a workflow designed to write data.

Related

Try it: the Formula Explainer: paste any formula and see each function and reference in plain English.

More in Formulas & functions

Sources: Function behaviour and menu paths: Microsoft Support documentation for Excel. Method & sources

Spotted an error in this guide? Let us know.