stringr str_detect() in R: Find Pattern in Strings

The str_detect() function in stringr returns TRUE or FALSE for each string indicating whether a pattern was found. It is the modern, vectorized replacement for base R grepl().

By Selva Prabhakaran · Published May 11, 2026 · Last updated May 11, 2026

⚡ Quick Answer

str_detect(x, "pattern")                            # regex (default)
str_detect(x, fixed("text"))                        # literal match
str_detect(x, regex("pattern", ignore_case = TRUE)) # case-insensitive
str_detect(x, "pattern", negate = TRUE)             # opposite (not match)
str_detect(c("apple","banana"), "an")               # vectorized
sum(str_detect(x, "pattern"))                       # count matches
filter(df, str_detect(name, "Smith"))               # use in dplyr filter

Need explanation? Read on for examples and pitfalls.

📊 Is str_detect() the right tool?

What str_detect() does in one sentence

str_detect(string, pattern) returns a logical vector of the same length as string, with TRUE where the pattern matches and FALSE elsewhere. The pattern is a regular expression by default; wrap with fixed() for literal string matching.

It is the workhorse for filtering data frames by string conditions, validating user input, and pre-processing text data.

Syntax

str_detect(string, pattern, negate = FALSE). Pattern is regex; use modifiers for fixed or case-insensitive.

Run live

Run live, no install needed. Every R block on this page runs in your browser. Click Run, edit the code, re-run instantly. No setup.

RLoad stringr and test pattern

library(stringr) library(tibble) x <- c("apple", "banana", "cherry", "Date") str_detect(x, "an") #> [1] TRUE TRUE FALSE FALSE

Tip

Default pattern is REGEX, so special characters need escaping. To match a literal ., use fixed(".") or \\.. To match a literal (, use \\(. The fixed() modifier is the simplest way to skip regex parsing entirely.

Five common patterns

1. Basic regex detection

RDoes each string contain 'an'?

str_detect(c("apple","banana","cherry"), "an") #> [1] TRUE TRUE FALSE

2. Literal (non-regex) match

RMatch a literal dot

str_detect(c("a.b", "ab", "a-b"), fixed(".")) #> [1] TRUE FALSE FALSE

fixed("text") treats text literally. Without fixed, . is regex for "any character".

3. Case-insensitive match

RFind 'apple' or 'Apple' or 'APPLE'

str_detect(c("apple","Apple","Banana"), regex("apple", ignore_case = TRUE)) #> [1] TRUE TRUE FALSE

regex(pattern, ignore_case = TRUE) is the case-insensitive variant.

4. Negate (find non-matches)

RStrings NOT containing 'an'

str_detect(c("apple","banana","cherry"), "an", negate = TRUE) #> [1] FALSE FALSE TRUE

negate = TRUE flips the result. Equivalent to !str_detect(...).

5. Use in dplyr filter

RFilter rows by string pattern

library(dplyr) df <- tibble::tibble(name = c("Alice Smith","Bob Jones","Carol Smith")) df |> filter(str_detect(name, "Smith")) #> # A tibble: 2 x 1 #> name #> <chr> #> 1 Alice Smith #> 2 Carol Smith

str_detect() returns a logical vector, perfect for filter().

Key Insight

str_detect() and grepl() produce the same results for simple patterns. stringr is more consistent (always returns the same length as input, NA in produces NA out) and reads cleaner in pipelines. For large vectors, performance is similar.

Common pitfalls

Pitfall 1: regex special chars treated literally. str_detect(x, "1.5") matches "1a5", "1b5", etc., not just "1.5". Use fixed("1.5") or escape: "1\\.5".

Pitfall 2: NA propagation. str_detect(NA, "x") returns NA, not FALSE. Filter with !is.na(x) & str_detect(...) to drop NAs cleanly.

Warning

Anchors ^ and $ match start/end of EACH STRING, not start/end of vector. str_detect(c("apple","banana"), "^a") returns c(TRUE, FALSE). To match whole strings exactly, use ^pattern$.

Try it yourself

Try it: Filter iris$Species (as a character vector) to rows where the species name contains "color". Save to ex_match.

RYour turn: detect species with 'color'

species <- as.character(iris$Species) ex_match <- # your code here unique(ex_match) #> Expected: "versicolor"

Click to reveal solution

RSolution

species <- as.character(iris$Species) ex_match <- species[str_detect(species, "color")] unique(ex_match) #> [1] "versicolor"

Explanation: str_detect(species, "color") returns TRUE for species containing "color" anywhere. species[...] subsets to those entries. Only "versicolor" matches.

After mastering str_detect, look at:

str_extract(): pull out the matched substring
str_replace(): replace matched pattern
str_count(): count matches per string
str_locate(): find position of match
str_subset(): return matching strings (filter + extract in one)
grepl(): base R equivalent

For complex patterns, the regex(), fixed(), coll(), and boundary() modifier functions in stringr give precise control over match behavior.

FAQ

How do I check if a string contains a pattern in R?

Use stringr::str_detect(string, "pattern"). Returns TRUE/FALSE per string. The pattern is a regex by default; wrap with fixed() for literal match.

What is the difference between str_detect and grepl in R?

Both return logical vectors. str_detect() is more consistent (NA in -> NA out, predictable lengths). grepl() is base R, no package needed. Choose stringr for tidyverse code; grepl for zero dependencies.

How do I do a case-insensitive str_detect in R?

Wrap the pattern in regex(pattern, ignore_case = TRUE): str_detect(x, regex("apple", ignore_case = TRUE)).

How do I check if a string starts with a substring in R?

Use anchor ^: str_detect(x, "^apple") returns TRUE for strings starting with "apple". Or use str_starts(x, "apple") for a clearer intent.

How do I count strings matching a pattern?

sum(str_detect(x, "pattern")) counts how many of the strings match. To count matches WITHIN each string, use str_count(x, "pattern") instead.

stringr str_detect() in R: Find Pattern in Strings

What str_detect() does in one sentence

Syntax

Five common patterns

1. Basic regex detection

2. Literal (non-regex) match

3. Case-insensitive match

4. Negate (find non-matches)

5. Use in dplyr filter

Common pitfalls

Try it yourself

Related stringr functions

FAQ