diff --git a/dtool/src/dtoolbase/pstrtod.cxx b/dtool/src/dtoolbase/pstrtod.cxx
index bed73a1b82..74e1d750df 100644
--- a/dtool/src/dtoolbase/pstrtod.cxx
+++ b/dtool/src/dtoolbase/pstrtod.cxx
@@ -17,31 +17,6 @@
 #include <ctype.h>
 #include <math.h>
 
-////////////////////////////////////////////////////////////////////
-//     Function: pstrnicmp
-//  Description: This case-insenstive string compare function is used
-//               by pstrtod, below.
-////////////////////////////////////////////////////////////////////
-static int
-pstrnicmp(const char *s1, const char *s2, size_t n) {
-  if (n == 0) {
-    // Trivial equivalence.
-    return 0;
-  }
-  while (tolower(*s1) == tolower(*s2)) {
-    --n;
-    if ((*s1) == '\0' || n == 0) {
-      // They reached the end of the string together.
-      return 0;
-    }
-    ++s1;
-    ++s2;
-  }
-
-  // There's a difference somewhere.
-  return (int)(unsigned char)(*s1) - (int)(unsigned char)(*s2);
-}
-
 
 ////////////////////////////////////////////////////////////////////
 //     Function: pstrtod
@@ -68,91 +43,65 @@ pstrtod(const char *nptr, char **endptr) {
   }
 
   double value = 0.0;
-  
-  if (pstrnicmp(p, "infinity", 8) == 0) {
-    value = INFINITY;
-    p += 8;
-  } else if (pstrnicmp(p, "inf", 3) == 0) {
-    value = INFINITY;
-    p += 3;
-  } else if (pstrnicmp(p, "nan", 3) == 0) {
-    value = NAN;
-    p += 3;
-    if (pstrnicmp(p, "0x", 2) == 0) {
-      // The NaN code also appears in the string.  Use it.
-      static const size_t nan_code_size = 32;
-      char nan_code[nan_code_size + 1];
-      nan_code[0] = '0';
-      nan_code[1] = 'x';
-      const char *q = p + 2;
-      int ni = 2;
-      while (isdigit(*q) || (tolower(*q) >= 'a' && tolower(*q) <= 'f')) {
-        if (ni < nan_code_size) {
-          nan_code[ni] = *q;
-          ++ni;
-        }
-        ++q;
-      }
-      nan_code[ni] = '\0';
-      value = nan(nan_code);
-      p = q;
-    }
 
-  } else {
-
-    // Start reading decimal digits to the left of the decimal point.
-    bool found_digits = false;
-    while (isdigit(*p)) {
-      value = (value * 10.0) + (*p - '0');
-      found_digits = true;
-      ++p;
-    }
-
-    if (*p == '.') {
-      ++p;
-      // Read decimal digits to the right of the decimal point.
-      double multiplicand = 0.1;
-      while (isdigit(*p)) {
-        value += (*p - '0') * multiplicand;
-        ++p;
-        found_digits = true;
-        multiplicand *= 0.1;
-      }
-    }
-
-    if (!found_digits) {
-      // Not a valid float.
-      if (endptr != NULL) {
-        *endptr = (char *)nptr;
-      }
-      return 0.0;
-    }
-
-    if (tolower(*p) == 'e') {
-      // There's an exponent.
-      ++p;
-
-      char esign = '+';
-      if (*p == '+' || *p == '-') {
-        esign = *p;
-        ++p;
-      }
-
-      // Start reading decimal digits to the left of the decimal point.
-      double evalue = 0.0;
-      while (isdigit(*p)) {
-        evalue = (evalue * 10.0) + (*p - '0');
-        ++p;
-      }
-
-      if (esign == '-') {
-        value /= pow(evalue, 10.0);
-      } else {
-        value *= pow(evalue, 10.0);
-      }
-    }        
+  if (isalpha(*p)) {
+    // For special cases like "inf" and "nan", pass these up to the
+    // system implementation of strtod.
+    return strtod(nptr, endptr);
   }
 
+  // Start reading decimal digits to the left of the decimal point.
+  bool found_digits = false;
+  while (isdigit(*p)) {
+    value = (value * 10.0) + (*p - '0');
+    found_digits = true;
+    ++p;
+  }
+  
+  if (*p == '.') {
+    ++p;
+    // Read decimal digits to the right of the decimal point.
+    double multiplicand = 0.1;
+    while (isdigit(*p)) {
+      value += (*p - '0') * multiplicand;
+      ++p;
+      found_digits = true;
+      multiplicand *= 0.1;
+    }
+  }
+  
+  if (!found_digits) {
+    // Not a valid float.
+    if (endptr != NULL) {
+      *endptr = (char *)nptr;
+    }
+    return 0.0;
+  }
+  
+  if (tolower(*p) == 'e') {
+    // There's an exponent.
+    ++p;
+    
+    char esign = '+';
+    if (*p == '+' || *p == '-') {
+      esign = *p;
+      ++p;
+    }
+    
+    // Start reading decimal digits to the left of the decimal point.
+    double evalue = 0.0;
+    while (isdigit(*p)) {
+      evalue = (evalue * 10.0) + (*p - '0');
+      ++p;
+    }
+    
+    if (esign == '-') {
+      value /= pow(evalue, 10.0);
+    } else {
+      value *= pow(evalue, 10.0);
+    }
+  }        
+
   if (sign == '-') {
     value = -value;
   }