diff --git a/libraries/AP_RangeFinder/AP_RangeFinder_NMEA.cpp b/libraries/AP_RangeFinder/AP_RangeFinder_NMEA.cpp
index 37788b968a..a15d925e0e 100644
--- a/libraries/AP_RangeFinder/AP_RangeFinder_NMEA.cpp
+++ b/libraries/AP_RangeFinder/AP_RangeFinder_NMEA.cpp
@@ -62,6 +62,10 @@ bool AP_RangeFinder_NMEA::decode(char c)
     case '\n':
     case '*':
     {
+        if (_sentence_done) {
+            return false;
+        }
+
         // null terminate and decode latest term
         _term[_term_offset] = 0;
         bool valid_sentence = decode_latest_term();
@@ -80,6 +84,7 @@ bool AP_RangeFinder_NMEA::decode(char c)
         _checksum = 0;
         _term_is_checksum = false;
         _distance_m = -1.0f;
+        _sentence_done = false;
         return false;
     }
 
@@ -100,6 +105,7 @@ bool AP_RangeFinder_NMEA::decode_latest_term()
 {
     // handle the last term in a message
     if (_term_is_checksum) {
+        _sentence_done = true;
         uint8_t nibble_high = 0;
         uint8_t nibble_low  = 0;
         if (!hex_to_uint8(_term[0], nibble_high) || !hex_to_uint8(_term[1], nibble_low)) {
diff --git a/libraries/AP_RangeFinder/AP_RangeFinder_NMEA.h b/libraries/AP_RangeFinder/AP_RangeFinder_NMEA.h
index 74b95eaa4e..81d1636fed 100644
--- a/libraries/AP_RangeFinder/AP_RangeFinder_NMEA.h
+++ b/libraries/AP_RangeFinder/AP_RangeFinder_NMEA.h
@@ -58,8 +58,9 @@ private:
     char _term[15];                         // buffer for the current term within the current sentence
     uint8_t _term_offset;                   // offset within the _term buffer where the next character should be placed
     uint8_t _term_number;                   // term index within the current sentence
-    float _distance_m = -1.0f;                      // distance in meters parsed from a term, -1 if no distance
+    float _distance_m = -1.0f;              // distance in meters parsed from a term, -1 if no distance
     uint8_t _checksum;                      // checksum accumulator
     bool _term_is_checksum;                 // current term is the checksum
     sentence_types _sentence_type;          // the sentence type currently being processed
+    bool _sentence_done;                    // true if this sentence has already been decoded
 };