diff --git a/CHANGELOG.md b/CHANGELOG.md
index d253a01eddfe..72db61edbd74 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -18,6 +18,7 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 - Fix incorrectly generated CSS when using square brackets inside arbitrary properties ([#11709](https://github.com/tailwindlabs/tailwindcss/pull/11709))
 - Make `content` optional for presets in TypeScript types ([#11730](https://github.com/tailwindlabs/tailwindcss/pull/11730))
 - Handle variable colors that have variable fallback values ([#12049](https://github.com/tailwindlabs/tailwindcss/pull/12049))
+- Batch reading content files to prevent `too many open files` error ([#12079](https://github.com/tailwindlabs/tailwindcss/pull/12079))
 
 ### Added
 
diff --git a/oxide/crates/core/src/lib.rs b/oxide/crates/core/src/lib.rs
index 9596c68556f7..c0bd47666baa 100644
--- a/oxide/crates/core/src/lib.rs
+++ b/oxide/crates/core/src/lib.rs
@@ -342,8 +342,8 @@ pub fn parse_candidate_strings(input: Vec<ChangedContent>, options: u8) -> Vec<S
 
     match (IO::from(options), Parsing::from(options)) {
         (IO::Sequential, Parsing::Sequential) => parse_all_blobs_sync(read_all_files_sync(input)),
-        (IO::Sequential, Parsing::Parallel) => parse_all_blobs_sync(read_all_files(input)),
-        (IO::Parallel, Parsing::Sequential) => parse_all_blobs(read_all_files_sync(input)),
+        (IO::Sequential, Parsing::Parallel) => parse_all_blobs(read_all_files_sync(input)),
+        (IO::Parallel, Parsing::Sequential) => parse_all_blobs_sync(read_all_files(input)),
         (IO::Parallel, Parsing::Parallel) => parse_all_blobs(read_all_files(input)),
     }
 }
diff --git a/src/lib/expandTailwindAtRules.js b/src/lib/expandTailwindAtRules.js
index b9b54bd6219e..7b4afaa8e1d8 100644
--- a/src/lib/expandTailwindAtRules.js
+++ b/src/lib/expandTailwindAtRules.js
@@ -135,43 +135,47 @@ export default function expandTailwindAtRules(context) {
 
     env.DEBUG && console.time('Reading changed files')
 
-    if (flagEnabled(context.tailwindConfig, 'oxideParser')) {
-      let rustParserContent = []
-      let regexParserContent = []
-
-      for (let item of context.changedContent) {
-        let transformer = getTransformer(context.tailwindConfig, item.extension)
-        let extractor = getExtractor(context, item.extension)
-
-        if (transformer === builtInTransformers.DEFAULT && extractor?.DEFAULT_EXTRACTOR === true) {
-          rustParserContent.push(item)
-        } else {
-          regexParserContent.push([item, { transformer, extractor }])
-        }
+    /** @type {[item: {file?: string, content?: string}, meta: {transformer: any, extractor: any}][]} */
+    let regexParserContent = []
+
+    /** @type {{file?: string, content?: string}[]} */
+    let rustParserContent = []
+
+    for (let item of context.changedContent) {
+      let transformer = getTransformer(context.tailwindConfig, item.extension)
+      let extractor = getExtractor(context, item.extension)
+
+      if (
+        flagEnabled(context.tailwindConfig, 'oxideParser') &&
+        transformer === builtInTransformers.DEFAULT &&
+        extractor?.DEFAULT_EXTRACTOR === true
+      ) {
+        rustParserContent.push(item)
+      } else {
+        regexParserContent.push([item, { transformer, extractor }])
       }
+    }
 
-      if (rustParserContent.length > 0) {
-        for (let candidate of parseCandidateStrings(
-          rustParserContent,
-          IO.Parallel | Parsing.Parallel
-        )) {
-          candidates.add(candidate)
-        }
+    // Read files using our newer, faster parser when:
+    // - Oxide is enabled; AND
+    // - The file is using default transfomers and extractors
+    if (rustParserContent.length > 0) {
+      for (let candidate of parseCandidateStrings(
+        rustParserContent,
+        IO.Parallel | Parsing.Parallel
+      )) {
+        candidates.add(candidate)
       }
+    }
+
+    // Otherwise, read any files in node and parse with regexes
+    const BATCH_SIZE = 500
+
+    for (let i = 0; i < regexParserContent.length; i += BATCH_SIZE) {
+      let batch = regexParserContent.slice(i, i + BATCH_SIZE)
 
-      if (regexParserContent.length > 0) {
-        await Promise.all(
-          regexParserContent.map(async ([{ file, content }, { transformer, extractor }]) => {
-            content = file ? await fs.promises.readFile(file, 'utf8') : content
-            getClassCandidates(transformer(content), extractor, candidates, seen)
-          })
-        )
-      }
-    } else {
       await Promise.all(
-        context.changedContent.map(async ({ file, content, extension }) => {
-          let transformer = getTransformer(context.tailwindConfig, extension)
-          let extractor = getExtractor(context, extension)
+        batch.map(async ([{ file, content }, { transformer, extractor }]) => {
           content = file ? await fs.promises.readFile(file, 'utf8') : content
           getClassCandidates(transformer(content), extractor, candidates, seen)
         })