diff --git a/openmp/libomptarget/plugins/amdgpu/src/rtl.cpp b/openmp/libomptarget/plugins/amdgpu/src/rtl.cpp
--- a/openmp/libomptarget/plugins/amdgpu/src/rtl.cpp
+++ b/openmp/libomptarget/plugins/amdgpu/src/rtl.cpp
@@ -1104,18 +1104,16 @@
        DeviceInfo.GroupsPerDevice[device_id]);
   }
 
-  // Adjust threads to the env variables
-  if (DeviceInfo.EnvTeamThreadLimit > 0 &&
-      (enforce_upper_bound(&DeviceInfo.NumThreads[device_id],
-                           DeviceInfo.EnvTeamThreadLimit))) {
-    DP("Capping max number of threads to OMP_TEAMS_THREAD_LIMIT=%d\n",
+  if (DeviceInfo.EnvTeamThreadLimit > 0) {
+    DeviceInfo.NumThreads[device_id] = DeviceInfo.EnvTeamThreadLimit;
+    DP("Number of threads set according to OMP_TEAMS_THREAD_LIMIT=%d\n",
        DeviceInfo.EnvTeamThreadLimit);
+  } else {
+    DeviceInfo.NumThreads[device_id] = RTLDeviceInfoTy::Default_WG_Size;
+    DP("Number of threads set according to library's default %d\n",
+       RTLDeviceInfoTy::Default_WG_Size);
   }
 
-  // Set default number of threads
-  DeviceInfo.NumThreads[device_id] = RTLDeviceInfoTy::Default_WG_Size;
-  DP("Default number of threads set according to library's default %d\n",
-     RTLDeviceInfoTy::Default_WG_Size);
   if (enforce_upper_bound(&DeviceInfo.NumThreads[device_id],
                           DeviceInfo.ThreadsPerGroup[device_id])) {
     DP("Default number of threads exceeds device limit, capping at %d\n",
@@ -2100,7 +2098,7 @@
    */
   int num_groups = 0;
 
-  int threadsPerGroup = RTLDeviceInfoTy::Default_WG_Size;
+  int threadsPerGroup = DeviceInfo.NumThreads[KernelInfo->device_id];
 
   getLaunchVals(threadsPerGroup, num_groups, KernelInfo->ConstWGSize,
                 KernelInfo->ExecutionMode, DeviceInfo.EnvTeamLimit,