diff --git a/src/libmsc/gsm_04_08_cc.c b/src/libmsc/gsm_04_08_cc.c
index e65fecbcf..cfa511742 100644
--- a/src/libmsc/gsm_04_08_cc.c
+++ b/src/libmsc/gsm_04_08_cc.c
@@ -768,6 +768,19 @@ void gsm48_cc_rx_setup_cn_local_rtp_port_known(struct gsm_trans *trans)
 	mncc_recvmsg(trans->net, trans, MNCC_SETUP_IND, &setup);
 }
 
+static bool codecs_match(const struct sdp_audio_codec *a, const struct sdp_audio_codec *b)
+{
+	if (!sdp_audio_codec_cmp(a, b, true, false))
+		return true;
+
+	/* Codecs are not identical. But will the MGW be able to transcode? */
+	if (!strcmp("AMR", a->subtype_name) && !strcmp("AMR", b->subtype_name)) {
+		/* TODO: compare matching mode-set. */
+		return true;
+	}
+	return false;
+}
+
 static void rx_mncc_sdp(struct gsm_trans *trans, uint32_t mncc_msg_type, const char *sdp,
 			const struct gsm_mncc_bearer_cap *bcap)
 {
@@ -814,15 +827,14 @@ static void rx_mncc_sdp(struct gsm_trans *trans, uint32_t mncc_msg_type, const c
 
 	/* Assume that the remote side put its actually assigned codec in first place. If our assigned codec doesn't
 	 * match that, continue re-assignment below. If they do match, exit here to leave as-is.
-	 * I'm not sure if this is common practice, or just a workaround: we cannot dynamically switch between all of
-	 * the payload types, but we still want to communicate all supported codecs. Maybe the proper way to tell the
-	 * other side to use a single codec is to leave only that single codec in the SDP; but at least MO *must* send
-	 * all possible codecs along, to give MT a chance to find a good match. Maybe MO should offer all supported
-	 * codecs, but MT should limit to a single pick -- but what do other SDP peers do, say via SIP? */
-	if (!sdp_audio_codec_cmp(&trans->cc.remote.audio_codecs.codec[0], &codecs->assignment, true, false)) {
-		/* remote's first codec matches our assigned codec. All is well. */
+	 * I'm not sure if assuming the special role for the codec that is listed first is common practice, or just a
+	 * workaround: we cannot dynamically switch between all of the payload types, but we still want to communicate
+	 * all supported codecs. Maybe the proper way to tell the other side to use a single codec is to leave only that
+	 * single codec in the SDP; but at least MO *must* send all possible codecs along, to give MT a chance to find a
+	 * good match. Maybe MO should offer all supported codecs, but MT should limit to a single pick -- but what do
+	 * other SDP peers do, say via SIP? */
+	if (codecs_match(&trans->cc.remote.audio_codecs.codec[0], &codecs->assignment))
 		return;
-	}
 
 	/* We've already completed Assignment of a voice channel (some time ago), and now the remote side has changed
 	 * to a mismatching codec (list). Try to re-assign this side to a matching codec. */
diff --git a/tests/msc_vlr/msc_vlr_test_call.c b/tests/msc_vlr/msc_vlr_test_call.c
index b959bc1c4..97be00adf 100644
--- a/tests/msc_vlr/msc_vlr_test_call.c
+++ b/tests/msc_vlr/msc_vlr_test_call.c
@@ -1258,10 +1258,12 @@ static const struct codec_test codec_tests[] = {
 		.mt_tx_sdp_mncc_alert_ind = SDP_CODECS_ALL_GSM_AMR_HR_FIRST,
 		/* mo_rx_sdp_mncc_alert_req == mt_tx_sdp_mncc_alert_ind */
 
+		/*
 		.mo_expect_reassignment = true,
 		.mo_tx_reassignment_perm_speech = PERM_SPEECH_ALL_GSM_HR3_FIRST,
 		.mo_rx_reassigned_codec_fr = false,
 		.mo_rx_reassigned_codec = "AMR:octet-align=1;mode-set=0,2,4",
+		*/
 
 		.mt_tx_sdp_mncc_setup_cnf = SDP_CODECS_ALL_GSM_AMR_HR_FIRST,
 		.mo_tx_sdp_mncc_setup_compl_ind = {},