diff --git a/travis/test/floatb.asm b/travis/test/floatb.asm
new file mode 100644
index 00000000..cbf9681b
--- /dev/null
+++ b/travis/test/floatb.asm
@@ -0,0 +1,32 @@
+	;; Known problematic floating-point numbers and their proper
+	;; encoding...
+
+	bits 64
+
+	dd 1.1e10
+	dd 0x5023e9ac	; Should be...
+
+	dd 50.40e9
+	dd 0x513bc130	; Should be...
+
+	dq 1.4e23
+	dq 0x44bda56a4b0835c0
+
+	dq 50.48e21
+	dq 0x44a5610d7502feae
+
+	dt 1.2e28
+	dq 0x9b18ab5df7180b6c
+	dw 0x405c
+
+	dt 50.46e25
+	dq 0xd0b29a67e95dcb60
+	dw 0x4057
+
+	;; Way too big numbers, should overflow to +Inf
+	dd 1.0E646456955
+	dd 1.0E646456956
+	dd 1.0E2147483646
+	dd 1.0E2147483647
+	dd 1.0E2147483648
+	dd 1.0E2147483649
diff --git a/travis/test/floatb.bin.t b/travis/test/floatb.bin.t
new file mode 100644
index 00000000..3fc9be23
Binary files /dev/null and b/travis/test/floatb.bin.t differ
diff --git a/travis/test/floatb.json b/travis/test/floatb.json
new file mode 100644
index 00000000..18ba8b6f
--- /dev/null
+++ b/travis/test/floatb.json
@@ -0,0 +1,12 @@
+[
+	{
+		"description": "Test of known floating-point bugs",
+		"id": "floatb",
+		"format": "bin",
+		"source": "floatb.asm",
+		"option": "-Ox",
+		"target": [
+			{ "output": "floatb.bin" }
+		]
+	}
+]