diff --git a/doc/src/Errors_messages.txt b/doc/src/Errors_messages.txt
index 4f3bbe8c24..b930dbce30 100644
--- a/doc/src/Errors_messages.txt
+++ b/doc/src/Errors_messages.txt
@@ -5828,6 +5828,12 @@ Must have periodic x,y dimensions and non-periodic z dimension to use
 Must have periodic x,y dimensions and non-periodic z dimension to use
 2d slab option with pppm/disp. :dd
 
+{Incorrect conversion in format string} :dt
+
+A format style variable was not using either a %f or a %g conversion.
+Or an immediate variable with format suffix was not using either
+a %f or a %g conversion in the format suffix. :dd
+
 {Incorrect element names in ADP potential file} :dt
 
 The element names in the ADP file do not match those requested. :dd
diff --git a/src/input.cpp b/src/input.cpp
index f88c8ca0c0..029c155ee8 100644
--- a/src/input.cpp
+++ b/src/input.cpp
@@ -50,6 +50,7 @@
 #include "accelerator_kokkos.h"
 #include "error.h"
 #include "memory.h"
+#include "utils.h"
 
 #ifdef _OPENMP
 #include <omp.h>
@@ -526,6 +527,11 @@ void Input::substitute(char *&str, char *&str2, int &max, int &max2, int flag)
           *fmtflag='\0';
         }
 
+        // quick check for proper format string
+
+        if (!utils::strmatch(fmtstr,"%[0-9 ]*\\.[0-9]+[fg]"))
+          error->all(FLERR,"Incorrect conversion in format string");
+
         snprintf(immediate,256,fmtstr,variable->compute_equal(var));
         value = immediate;
 
diff --git a/src/input.h b/src/input.h
index 47ad7779f1..d62ab11234 100644
--- a/src/input.h
+++ b/src/input.h
@@ -173,6 +173,11 @@ E: Unbalanced quotes in input line
 No matching end double quote was found following a leading double
 quote.
 
+E: Incorrect conversion in format string
+
+An immediate variable with format suffix was not using
+either a %f or a %g conversion in the format suffix.
+
 E: Input line quote not followed by white-space
 
 An end quote must be followed by white-space.
diff --git a/src/variable.cpp b/src/variable.cpp
index 56b66cad0d..0fe18700ee 100644
--- a/src/variable.cpp
+++ b/src/variable.cpp
@@ -38,6 +38,7 @@
 #include "memory.h"
 #include "info.h"
 #include "error.h"
+#include "utils.h"
 
 using namespace LAMMPS_NS;
 using namespace MathConst;
@@ -383,6 +384,8 @@ void Variable::set(int narg, char **arg)
     num[nvar] = 3;
     which[nvar] = 0;
     pad[nvar] = 0;
+    if (!utils::strmatch(arg[2],"%[0-9 ]*\\.[0-9]+[fg]"))
+      error->all(FLERR,"Incorrect conversion in format string");
     data[nvar] = new char*[num[nvar]];
     copy(2,&arg[2],data[nvar]);
     data[nvar][2] = new char[VALUELENGTH];
diff --git a/src/variable.h b/src/variable.h
index a37ee4cff7..a071fdb021 100644
--- a/src/variable.h
+++ b/src/variable.h
@@ -210,6 +210,10 @@ E: Invalid variable style with next command
 Variable styles {equal} and {world} cannot be used in a next
 command.
 
+E: Incorrect conversion in format string
+
+A format style variable was not using either a %f or a %g conversion.
+
 E: Next command must list all universe and uloop variables
 
 This is to insure they stay in sync.