diff --git a/CHANGELOG.md b/CHANGELOG.md
index 2e82be4d0fe..06e50a92d6a 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -25,6 +25,8 @@
 
     This mirrors how the JS code generator similarly avoids the character sequence `</script`.
 
+    In addition, the check that escapes `</style` and `</script` is now case-insensitive to match how the browser's HTML parser behaves. So `</STYLE` and `</SCRIPT` are now escaped as well.
+
 * Fix a TypeScript parsing edge case with ASI (Automatic Semicolon Insertion) ([#1512](https://github.com/evanw/esbuild/issues/1512))
 
     This fixes a parsing bug where TypeScript types consisting of multiple identifiers joined together with a `.` could incorrectly extend onto the next line if the next line started with `<`. This problem was due to ASI; esbuild should be automatically inserting a semicolon at the end of the line:
diff --git a/internal/css_printer/css_printer.go b/internal/css_printer/css_printer.go
index 468554d904a..aa592134720 100644
--- a/internal/css_printer/css_printer.go
+++ b/internal/css_printer/css_printer.go
@@ -488,7 +488,7 @@ func (p *printer) printQuotedWithQuote(text string, quote byte) {
 
 		case '/':
 			// Avoid generating the sequence "</style" in CSS code
-			if i >= 1 && text[i-1] == '<' && strings.HasPrefix(text[i+1:], "style") {
+			if i >= 1 && text[i-1] == '<' && i+6 <= len(text) && strings.EqualFold(text[i+1:i+6], "style") {
 				escape = escapeBackslash
 			}
 
diff --git a/internal/css_printer/css_printer_test.go b/internal/css_printer/css_printer_test.go
index d1964e890a8..7fde4dc06f3 100644
--- a/internal/css_printer/css_printer_test.go
+++ b/internal/css_printer/css_printer_test.go
@@ -90,7 +90,11 @@ func TestStringQuote(t *testing.T) {
 	expectPrintedString(t, "</script>", "\"</script>\"")
 	expectPrintedString(t, "</style>", "\"<\\/style>\"")
 	expectPrintedString(t, "</style", "\"<\\/style\"")
+	expectPrintedString(t, "</STYLE", "\"<\\/STYLE\"")
+	expectPrintedString(t, "</StYlE", "\"<\\/StYlE\"")
 	expectPrintedString(t, ">/style", "\">/style\"")
+	expectPrintedString(t, ">/STYLE", "\">/STYLE\"")
+	expectPrintedString(t, ">/StYlE", "\">/StYlE\"")
 }
 
 func TestURLQuote(t *testing.T) {
diff --git a/internal/js_lexer/js_lexer.go b/internal/js_lexer/js_lexer.go
index 9257ff7aae1..ed161e26aab 100644
--- a/internal/js_lexer/js_lexer.go
+++ b/internal/js_lexer/js_lexer.go
@@ -2840,7 +2840,7 @@ func StringToUTF16(text string) []uint16 {
 }
 
 func UTF16ToString(text []uint16) string {
-	temp := make([]byte, utf8.UTFMax)
+	var temp [utf8.UTFMax]byte
 	b := strings.Builder{}
 	n := len(text)
 	for i := 0; i < n; i++ {
@@ -2851,14 +2851,14 @@ func UTF16ToString(text []uint16) string {
 				i++
 			}
 		}
-		width := encodeWTF8Rune(temp, r1)
+		width := encodeWTF8Rune(temp[:], r1)
 		b.Write(temp[:width])
 	}
 	return b.String()
 }
 
 func UTF16ToStringWithValidation(text []uint16) (string, uint16, bool) {
-	temp := make([]byte, utf8.UTFMax)
+	var temp [utf8.UTFMax]byte
 	b := strings.Builder{}
 	n := len(text)
 	for i := 0; i < n; i++ {
@@ -2877,7 +2877,7 @@ func UTF16ToStringWithValidation(text []uint16) (string, uint16, bool) {
 		} else if r1 >= 0xDC00 && r1 <= 0xDFFF {
 			return "", uint16(r1), false
 		}
-		width := encodeWTF8Rune(temp, r1)
+		width := encodeWTF8Rune(temp[:], r1)
 		b.Write(temp[:width])
 	}
 	return b.String(), 0, true
@@ -2889,7 +2889,7 @@ func UTF16EqualsString(text []uint16, str string) bool {
 		// Strings can't be equal if UTF-16 encoding is longer than UTF-8 encoding
 		return false
 	}
-	temp := [utf8.UTFMax]byte{}
+	var temp [utf8.UTFMax]byte
 	n := len(text)
 	j := 0
 	for i := 0; i < n; i++ {
diff --git a/internal/js_parser/js_parser.go b/internal/js_parser/js_parser.go
index fa061324232..c1a3fe815ef 100644
--- a/internal/js_parser/js_parser.go
+++ b/internal/js_parser/js_parser.go
@@ -10515,6 +10515,20 @@ func (p *parser) mangleTemplate(loc logger.Loc, e *js_ast.ETemplate) js_ast.Expr
 	return js_ast.Expr{Loc: loc, Data: e}
 }
 
+func containsClosingScriptTag(text string) bool {
+	for {
+		i := strings.Index(text, "</")
+		if i < 0 {
+			break
+		}
+		text = text[i+2:]
+		if len(text) >= 6 && strings.EqualFold(text[:6], "script") {
+			return true
+		}
+	}
+	return false
+}
+
 // This function takes "exprIn" as input from the caller and produces "exprOut"
 // for the caller to pass along extra data. This is mostly for optional chaining.
 func (p *parser) visitExprInOut(expr js_ast.Expr, in exprIn) (js_ast.Expr, exprOut) {
@@ -10785,11 +10799,11 @@ func (p *parser) visitExprInOut(expr js_ast.Expr, in exprIn) (js_ast.Expr, exprO
 		// Lower tagged template literals that include "</script"
 		// since we won't be able to escape it without lowering it
 		if !shouldLowerTemplateLiteral && e.TagOrNil.Data != nil {
-			if strings.Contains(e.HeadRaw, "</script") {
+			if containsClosingScriptTag(e.HeadRaw) {
 				shouldLowerTemplateLiteral = true
 			} else {
 				for _, part := range e.Parts {
-					if strings.Contains(part.TailRaw, "</script") {
+					if containsClosingScriptTag(part.TailRaw) {
 						shouldLowerTemplateLiteral = true
 						break
 					}
diff --git a/internal/js_printer/js_printer.go b/internal/js_printer/js_printer.go
index 9569943b6b9..052eb543768 100644
--- a/internal/js_printer/js_printer.go
+++ b/internal/js_printer/js_printer.go
@@ -217,8 +217,23 @@ func (p *printer) printUnquotedUTF16(text []uint16, quote rune) {
 
 		case '/':
 			// Avoid generating the sequence "</script" in JS code
-			if i >= 2 && text[i-2] == '<' && i+6 <= len(text) && js_lexer.UTF16EqualsString(text[i:i+6], "script") {
-				js = append(js, '\\')
+			if i >= 2 && text[i-2] == '<' && i+6 <= len(text) {
+				script := "script"
+				matches := true
+				for j := 0; j < 6; j++ {
+					a := text[i+j]
+					b := uint16(script[j])
+					if a >= 'A' && a <= 'Z' {
+						a += 'a' - 'A'
+					}
+					if a != b {
+						matches = false
+						break
+					}
+				}
+				if matches {
+					js = append(js, '\\')
+				}
 			}
 			js = append(js, '/')
 
@@ -1781,7 +1796,7 @@ func (p *printer) printExpr(expr js_ast.Expr, level js_ast.L, flags printExprFla
 
 		if n > 0 {
 			// Avoid forming a single-line comment or "</script" sequence
-			if last := buffer[n-1]; last == '/' || (last == '<' && strings.HasPrefix(e.Value, "/script")) {
+			if last := buffer[n-1]; last == '/' || (last == '<' && len(e.Value) >= 7 && strings.EqualFold(e.Value[:7], "/script")) {
 				p.print(" ")
 			}
 		}
@@ -2405,9 +2420,30 @@ func (p *printer) printIf(s *js_ast.SIf) {
 	}
 }
 
+func escapeClosingScriptTag(text string) string {
+	i := strings.Index(text, "</")
+	if i < 0 {
+		return text
+	}
+	var b strings.Builder
+	for {
+		b.WriteString(text[:i+1])
+		text = text[i+1:]
+		if len(text) >= 7 && strings.EqualFold(text[:7], "/script") {
+			b.WriteByte('\\')
+		}
+		i = strings.Index(text, "</")
+		if i < 0 {
+			break
+		}
+	}
+	b.WriteString(text)
+	return b.String()
+}
+
 func (p *printer) printIndentedComment(text string) {
 	// Avoid generating a comment containing the character sequence "</script"
-	text = strings.ReplaceAll(text, "</script", "<\\/script")
+	text = escapeClosingScriptTag(text)
 
 	if strings.HasPrefix(text, "/*") {
 		// Re-indent multi-line comments
diff --git a/internal/js_printer/js_printer_test.go b/internal/js_printer/js_printer_test.go
index b711983d2a2..8ca9fab1ec7 100644
--- a/internal/js_printer/js_printer_test.go
+++ b/internal/js_printer/js_printer_test.go
@@ -910,17 +910,30 @@ func TestAvoidSlashScript(t *testing.T) {
 	// Positive cases
 	expectPrinted(t, "x = '</script'", "x = \"<\\/script\";\n")
 	expectPrinted(t, "x = `</script`", "x = `<\\/script`;\n")
+	expectPrinted(t, "x = `</SCRIPT`", "x = `<\\/SCRIPT`;\n")
+	expectPrinted(t, "x = `</ScRiPt`", "x = `<\\/ScRiPt`;\n")
 	expectPrinted(t, "x = `</script${y}`", "x = `<\\/script${y}`;\n")
 	expectPrinted(t, "x = `${y}</script`", "x = `${y}<\\/script`;\n")
 	expectPrintedMinify(t, "x = 1 < /script/.exec(y).length", "x=1< /script/.exec(y).length;")
+	expectPrintedMinify(t, "x = 1 < /SCRIPT/.exec(y).length", "x=1< /SCRIPT/.exec(y).length;")
+	expectPrintedMinify(t, "x = 1 < /ScRiPt/.exec(y).length", "x=1< /ScRiPt/.exec(y).length;")
 	expectPrintedMinify(t, "x = 1 << /script/.exec(y).length", "x=1<< /script/.exec(y).length;")
-	expectPrinted(t, "//! </script", "//! <\\/script\n")
+	expectPrinted(t, "//! </script\n//! >/script\n//! /script", "//! <\\/script\n//! >/script\n//! /script\n")
+	expectPrinted(t, "//! </SCRIPT\n//! >/SCRIPT\n//! /SCRIPT", "//! <\\/SCRIPT\n//! >/SCRIPT\n//! /SCRIPT\n")
+	expectPrinted(t, "//! </ScRiPt\n//! >/ScRiPt\n//! /ScRiPt", "//! <\\/ScRiPt\n//! >/ScRiPt\n//! /ScRiPt\n")
+	expectPrinted(t, "/*! </script \n </script */", "/*! <\\/script \n <\\/script */\n")
+	expectPrinted(t, "/*! </SCRIPT \n </SCRIPT */", "/*! <\\/SCRIPT \n <\\/SCRIPT */\n")
+	expectPrinted(t, "/*! </ScRiPt \n </ScRiPt */", "/*! <\\/ScRiPt \n <\\/ScRiPt */\n")
 	expectPrinted(t, "String.raw`</script`",
 		"var _a;\nString.raw(_a || (_a = __template([\"<\\/script\"])));\nimport {\n  __template\n} from \"<runtime>\";\n")
 	expectPrinted(t, "String.raw`</script${a}`",
 		"var _a;\nString.raw(_a || (_a = __template([\"<\\/script\", \"\"])), a);\nimport {\n  __template\n} from \"<runtime>\";\n")
 	expectPrinted(t, "String.raw`${a}</script`",
 		"var _a;\nString.raw(_a || (_a = __template([\"\", \"<\\/script\"])), a);\nimport {\n  __template\n} from \"<runtime>\";\n")
+	expectPrinted(t, "String.raw`</SCRIPT`",
+		"var _a;\nString.raw(_a || (_a = __template([\"<\\/SCRIPT\"])));\nimport {\n  __template\n} from \"<runtime>\";\n")
+	expectPrinted(t, "String.raw`</ScRiPt`",
+		"var _a;\nString.raw(_a || (_a = __template([\"<\\/ScRiPt\"])));\nimport {\n  __template\n} from \"<runtime>\";\n")
 
 	// Negative cases
 	expectPrinted(t, "x = '</'", "x = \"</\";\n")