diff --git a/modules/markup/markdown/markdown_math_test.go b/modules/markup/markdown/markdown_math_test.go
index a75f18d36a..9e368cb689 100644
--- a/modules/markup/markdown/markdown_math_test.go
+++ b/modules/markup/markdown/markdown_math_test.go
@@ -30,6 +30,10 @@ func TestMathRender(t *testing.T) {
 			"$ a $",
 			`<p><code class="language-math">a</code></p>` + nl,
 		},
+		{
+			"$a$$b$",
+			`<p><code class="language-math">a</code><code class="language-math">b</code></p>` + nl,
+		},
 		{
 			"$a$ $b$",
 			`<p><code class="language-math">a</code> <code class="language-math">b</code></p>` + nl,
@@ -59,7 +63,7 @@ func TestMathRender(t *testing.T) {
 			`<p>a$b $a a$b b$</p>` + nl,
 		},
 		{
-			"a$x$",
+			"a$x$", // Pattern: "word$other$" The real world example is: "Price is between US$1 and US$2.", so don't parse this.
 			`<p>a$x$</p>` + nl,
 		},
 		{
@@ -70,6 +74,10 @@ func TestMathRender(t *testing.T) {
 			"$a$ ($b$) [$c$] {$d$}",
 			`<p><code class="language-math">a</code> (<code class="language-math">b</code>) [$c$] {$d$}</p>` + nl,
 		},
+		{
+			"[$a$](link)",
+			`<p><a href="/link" rel="nofollow"><code class="language-math">a</code></a></p>` + nl,
+		},
 		{
 			"$$a$$",
 			`<p><code class="language-math">a</code></p>` + nl,
diff --git a/modules/markup/markdown/math/inline_parser.go b/modules/markup/markdown/math/inline_parser.go
index a711d1e1cd..564861df90 100644
--- a/modules/markup/markdown/math/inline_parser.go
+++ b/modules/markup/markdown/math/inline_parser.go
@@ -54,6 +54,10 @@ func isAlphanumeric(b byte) bool {
 	return (b >= 'a' && b <= 'z') || (b >= 'A' && b <= 'Z') || (b >= '0' && b <= '9')
 }
 
+func isInMarkdownLinkText(block text.Reader, lineAfter []byte) bool {
+	return block.PrecendingCharacter() == '[' && bytes.HasPrefix(lineAfter, []byte("]("))
+}
+
 // Parse parses the current line and returns a result of parsing.
 func (parser *inlineParser) Parse(parent ast.Node, block text.Reader, pc parser.Context) ast.Node {
 	line, _ := block.PeekLine()
@@ -115,7 +119,9 @@ func (parser *inlineParser) Parse(parent ast.Node, block text.Reader, pc parser.
 			}
 			// check valid ending character
 			isValidEndingChar := isPunctuation(succeedingCharacter) || isParenthesesClose(succeedingCharacter) ||
-				succeedingCharacter == ' ' || succeedingCharacter == '\n' || succeedingCharacter == 0
+				succeedingCharacter == ' ' || succeedingCharacter == '\n' || succeedingCharacter == 0 ||
+				succeedingCharacter == '$' ||
+				isInMarkdownLinkText(block, line[i+len(stopMark):])
 			if checkSurrounding && !isValidEndingChar {
 				break
 			}