diff --git a/interface/src/avatar/MyAvatar.cpp b/interface/src/avatar/MyAvatar.cpp
index 6dc3d820ae..ab4989a651 100644
--- a/interface/src/avatar/MyAvatar.cpp
+++ b/interface/src/avatar/MyAvatar.cpp
@@ -1028,7 +1028,28 @@ void MyAvatar::renderBody(ViewFrustum* renderFrustum, RenderMode renderMode, boo
     if (!(_skeletonModel.isRenderable() && getHead()->getFaceModel().isRenderable())) {
         return; // wait until both models are loaded
     }
-    
+
+    Camera *camera = Application::getInstance()->getCamera();
+    const glm::vec3 cameraPos = camera->getPosition();
+
+    // Set near clip distance according to skeleton model dimensions if first person and there is no separate head model.
+    if (shouldRenderHead(cameraPos, renderMode) || !getHead()->getFaceModel().getURL().isEmpty()) {
+        renderFrustum->setNearClip(DEFAULT_NEAR_CLIP);
+    } else {
+        float clipDistance = _skeletonModel.getHeadClipDistance();
+        if (OculusManager::isConnected()) {
+            // If avatar is horizontally in front of camera, increase clip distance by the amount it is in front.
+            glm::vec3 cameraToAvatar = _position - cameraPos;
+            cameraToAvatar.y = 0.0f;
+            glm::vec3 cameraLookAt = camera->getOrientation() * glm::vec3(0.0f, 0.0f, -1.0f);
+            float headOffset = glm::dot(cameraLookAt, cameraToAvatar);
+            if (headOffset > 0) {
+                clipDistance += headOffset;
+            }
+        }
+        renderFrustum->setNearClip(clipDistance);
+    }
+
     //  Render the body's voxels and head
     Model::RenderMode modelRenderMode = (renderMode == SHADOW_RENDER_MODE) ?
         Model::SHADOW_RENDER_MODE : Model::DEFAULT_RENDER_MODE;
@@ -1040,8 +1061,6 @@ void MyAvatar::renderBody(ViewFrustum* renderFrustum, RenderMode renderMode, boo
     }
     
     //  Render head so long as the camera isn't inside it
-    const Camera *camera = Application::getInstance()->getCamera();
-    const glm::vec3 cameraPos = camera->getPosition();
     if (shouldRenderHead(cameraPos, renderMode)) {
         getHead()->render(1.0f, renderFrustum, modelRenderMode, postLighting);
     }
diff --git a/interface/src/avatar/SkeletonModel.cpp b/interface/src/avatar/SkeletonModel.cpp
index 4fdebd5f6f..3a61f69dee 100644
--- a/interface/src/avatar/SkeletonModel.cpp
+++ b/interface/src/avatar/SkeletonModel.cpp
@@ -37,7 +37,8 @@ SkeletonModel::SkeletonModel(Avatar* owningAvatar, QObject* parent) :
     _defaultEyeModelPosition(glm::vec3(0.0f, 0.0f, 0.0f)),
     _standingFoot(NO_FOOT),
     _standingOffset(0.0f),
-    _clampedFootPosition(0.0f)
+    _clampedFootPosition(0.0f),
+    _headClipDistance(DEFAULT_NEAR_CLIP)
 {
 }
 
@@ -78,6 +79,10 @@ void SkeletonModel::setJointStates(QVector<JointState> states) {
         buildShapes();
     }
 
+    Extents meshExtents = getMeshExtents();
+    _headClipDistance = -(meshExtents.minimum.z / _scale.z - _defaultEyeModelPosition.z);
+    _headClipDistance = std::max(_headClipDistance, DEFAULT_NEAR_CLIP);
+
     emit skeletonLoaded();
 }
 
diff --git a/interface/src/avatar/SkeletonModel.h b/interface/src/avatar/SkeletonModel.h
index 5427fcaf25..298d74fb7a 100644
--- a/interface/src/avatar/SkeletonModel.h
+++ b/interface/src/avatar/SkeletonModel.h
@@ -110,6 +110,8 @@ public:
 
     bool hasSkeleton();
 
+    float getHeadClipDistance() const { return _headClipDistance; }
+
 signals:
 
     void skeletonLoaded();
@@ -160,6 +162,8 @@ private:
     int _standingFoot;
     glm::vec3 _standingOffset;
     glm::vec3 _clampedFootPosition;
+
+    float _headClipDistance;  // Near clip distance to use if no separate head model
 };
 
 #endif // hifi_SkeletonModel_h