diff --git a/go.sum b/go.sum
index 9f5b5bea3..951e4c6f9 100644
--- a/go.sum
+++ b/go.sum
@@ -68,6 +68,7 @@ github.com/magefile/mage v1.4.0 h1:RI7B1CgnPAuu2O9lWszwya61RLmfL0KCdo+QyyI/Bhk=
 github.com/magefile/mage v1.4.0/go.mod h1:IUDi13rsHje59lecXokTfGX0QIzO45uVPlXnJYsXepA=
 github.com/magiconair/properties v1.8.0 h1:LLgXmsheXeRoUOBOjtwPQCWIYqM/LU1ayDtDePerRcY=
 github.com/magiconair/properties v1.8.0/go.mod h1:PppfXfuXeibc/6YijjN8zIbojt8czPbwD3XqdrwzmxQ=
+github.com/markbates/inflect v0.0.0-20171215194931-a12c3aec81a6 h1:LZhVjIISSbj8qLf2qDPP0D8z0uvOWAW5C85ly5mJW6c=
 github.com/markbates/inflect v0.0.0-20171215194931-a12c3aec81a6/go.mod h1:oTeZL2KHA7CUX6X+fovmK9OvIOFuqu0TwdQrZjLTh88=
 github.com/matryer/try v0.0.0-20161228173917-9ac251b645a2/go.mod h1:0KeJpeMD6o+O4hW7qJOT7vyQPKrWmj26uf5wMc/IiIs=
 github.com/mattn/go-isatty v0.0.4 h1:bnP0vzxcAdeI1zdubAl5PjU6zsERjGZb7raWodagDYs=
diff --git a/hugolib/page_content.go b/hugolib/page_content.go
index 75124a1df..a363249d5 100644
--- a/hugolib/page_content.go
+++ b/hugolib/page_content.go
@@ -83,8 +83,6 @@ Loop:
 
 		switch {
 		case it.Type == pageparser.TypeIgnore:
-		case it.Type == pageparser.TypeHTMLComment:
-			// Ignore. This is only a leading Front matter comment.
 		case it.Type == pageparser.TypeHTMLStart:
 			// This is HTML without front matter. It can still have shortcodes.
 			p.renderable = false
diff --git a/hugolib/page_test.go b/hugolib/page_test.go
index f9c26754b..cd8bd8ffc 100644
--- a/hugolib/page_test.go
+++ b/hugolib/page_test.go
@@ -1586,6 +1586,33 @@ CONTENT:{{ .Content }}
 	)
 }
 
+// https://github.com/gohugoio/hugo/issues/5478
+func TestPageWithCommentedOutFrontMatter(t *testing.T) {
+	b := newTestSitesBuilder(t)
+	b.WithSimpleConfigFile()
+
+	b.WithContent("page.md", `<!--
++++
+title = "hello"
++++
+-->
+This is the content.
+`)
+
+	b.WithTemplatesAdded("layouts/_default/single.html", `
+Title: {{ .Title }}
+Content:{{ .Content }}
+`)
+
+	b.CreateSites().Build(BuildCfg{})
+
+	b.AssertFileContent("public/page/index.html",
+		"Title: hello",
+		"Content:<p>This is the content.</p>",
+	)
+
+}
+
 // TODO(bep) this may be useful for other tests.
 func compareObjects(a interface{}, b interface{}) bool {
 	aStr := strings.Split(fmt.Sprintf("%v", a), "")
diff --git a/parser/pageparser/item.go b/parser/pageparser/item.go
index 644c20e87..b9d47b16e 100644
--- a/parser/pageparser/item.go
+++ b/parser/pageparser/item.go
@@ -108,7 +108,6 @@ const (
 
 	// page items
 	TypeHTMLStart          // document starting with < as first non-whitespace
-	TypeHTMLComment        // We ignore leading comments
 	TypeLeadSummaryDivider // <!--more-->,  # more
 	TypeFrontMatterYAML
 	TypeFrontMatterTOML
diff --git a/parser/pageparser/pagelexer.go b/parser/pageparser/pagelexer.go
index 94c1ff26b..5802c318b 100644
--- a/parser/pageparser/pagelexer.go
+++ b/parser/pageparser/pagelexer.go
@@ -53,6 +53,8 @@ type pageLexer struct {
 	summaryDivider []byte
 	// Set when we have parsed any summary divider
 	summaryDividerChecked bool
+	// Whether we're in a HTML comment.
+	isInHTMLComment bool
 
 	lexerShortcodeState
 
@@ -120,7 +122,7 @@ var (
 	delimYAML         = []byte("---")
 	delimOrg          = []byte("#+")
 	htmlCommentStart  = []byte("<!--")
-	htmlCOmmentEnd    = []byte("-->")
+	htmlCommentEnd    = []byte("-->")
 )
 
 func (l *pageLexer) next() rune {
@@ -195,6 +197,15 @@ func (l *pageLexer) consumeCRLF() bool {
 	return consumed
 }
 
+func (l *pageLexer) consumeToNextLine() {
+	for {
+		r := l.next()
+		if r == eof || isEndOfLine(r) {
+			return
+		}
+	}
+}
+
 func (l *pageLexer) consumeSpace() {
 	for {
 		r := l.next()
@@ -206,6 +217,10 @@ func (l *pageLexer) consumeSpace() {
 }
 
 func lexMainSection(l *pageLexer) stateFunc {
+	if l.isInHTMLComment {
+		return lexEndFromtMatterHTMLComment
+	}
+
 	// Fast forward as far as possible.
 	var l1, l2 int
 
@@ -312,16 +327,15 @@ LOOP:
 		case r == byteOrderMark:
 			l.emit(TypeIgnore)
 		case !isSpace(r) && !isEndOfLine(r):
-			// No front matter.
 			if r == '<' {
 				l.backup()
 				if l.hasPrefix(htmlCommentStart) {
-					right := l.index(htmlCOmmentEnd)
-					if right == -1 {
-						return l.errorf("starting HTML comment with no end")
-					}
-					l.pos += right + len(htmlCOmmentEnd)
-					l.emit(TypeHTMLComment)
+					// This may be commented out front mattter, which should
+					// still be read.
+					l.consumeToNextLine()
+					l.isInHTMLComment = true
+					l.emit(TypeIgnore)
+					continue LOOP
 				} else {
 					if l.pos > l.start {
 						l.emit(tText)
@@ -341,6 +355,19 @@ LOOP:
 	return lexMainSection
 }
 
+func lexEndFromtMatterHTMLComment(l *pageLexer) stateFunc {
+	l.isInHTMLComment = false
+	right := l.index(htmlCommentEnd)
+	if right == -1 {
+		return l.errorf("starting HTML comment with no end")
+	}
+	l.pos += right + len(htmlCommentEnd)
+	l.emit(TypeIgnore)
+
+	// Now move on to the shortcodes.
+	return lexMainSection
+}
+
 func lexDone(l *pageLexer) stateFunc {
 
 	// Done!
diff --git a/parser/pageparser/pageparser_intro_test.go b/parser/pageparser/pageparser_intro_test.go
index ba48a3ee3..901b216c8 100644
--- a/parser/pageparser/pageparser_intro_test.go
+++ b/parser/pageparser/pageparser_intro_test.go
@@ -60,7 +60,8 @@ var frontMatterTests = []lexerTest{
 	{"No front matter", "\nSome text.\n", []Item{tstSomeText, tstEOF}},
 	{"YAML front matter", "---\nfoo: \"bar\"\n---\n\nSome text.\n", []Item{tstFrontMatterYAML, tstSomeText, tstEOF}},
 	{"YAML empty front matter", "---\n---\n\nSome text.\n", []Item{nti(TypeFrontMatterYAML, ""), tstSomeText, tstEOF}},
-	{"YAML commented out front matter", "<!--\n---\nfoo: \"bar\"\n---\n-->\nSome text.\n", []Item{nti(TypeHTMLComment, "<!--\n---\nfoo: \"bar\"\n---\n-->"), tstSomeText, tstEOF}},
+	{"YAML commented out front matter", "<!--\n---\nfoo: \"bar\"\n---\n-->\nSome text.\n", []Item{nti(TypeIgnore, "<!--\n"), tstFrontMatterYAML, nti(TypeIgnore, "-->"), tstSomeText, tstEOF}},
+	{"YAML commented out front matter, no end", "<!--\n---\nfoo: \"bar\"\n---\nSome text.\n", []Item{nti(TypeIgnore, "<!--\n"), tstFrontMatterYAML, nti(tError, "starting HTML comment with no end")}},
 	// Note that we keep all bytes as they are, but we need to handle CRLF
 	{"YAML front matter CRLF", "---\r\nfoo: \"bar\"\r\n---\n\nSome text.\n", []Item{tstFrontMatterYAMLCRLF, tstSomeText, tstEOF}},
 	{"TOML front matter", "+++\nfoo = \"bar\"\n+++\n\nSome text.\n", []Item{tstFrontMatterTOML, tstSomeText, tstEOF}},
@@ -78,7 +79,6 @@ var frontMatterTests = []lexerTest{
 func TestFrontMatter(t *testing.T) {
 	t.Parallel()
 	for i, test := range frontMatterTests {
-
 		items := collect([]byte(test.input), false, lexIntroSection)
 		if !equal(items, test.items) {
 			got := crLfReplacer.Replace(fmt.Sprint(items))