diff --git a/doc/compute_rdf.html b/doc/compute_rdf.html
new file mode 100644
index 000000000..2fd33f36b
--- /dev/null
+++ b/doc/compute_rdf.html
@@ -0,0 +1,130 @@
+<HTML>
+<CENTER><A HREF = "http://lammps.sandia.gov">LAMMPS WWW Site</A> - <A HREF = "Manual.html">LAMMPS Documentation</A> - <A HREF = "Section_commands.html#comm">LAMMPS Commands</A>
+</CENTER>
+
+
+
+
+
+
+<HR>
+
+<H3>compute rdf command
+</H3>
+<P><B>Syntax:</B>
+</P>
+<PRE>compute ID group-ID rdf Nbin itype1 jtype1 itype2 jtype2 ...
+</PRE>
+<UL><LI>ID, group-ID are documented in <A HREF = "compute.html">compute</A> command
+<LI>rdf = style name of this compute command
+<LI>Nbin = number of RDF bins
+<LI>itypeN = central atom type for Nth RDF histogram (see asterisk form below)
+<LI>jtypeN = distribution atom type for Nth RDF histogram (see asterisk form below)
+</UL>
+<P><B>Examples:</B>
+</P>
+<PRE>fix 1 all rdf 100
+fix 1 all rdf 100 1 1
+fix 1 all rdf 100 * 3
+fix 1 fluid rdf 500 1 1 1 2 2 1 2 2
+fix 1 fluid rdf 500 1*3 2 5 *10
+</PRE>
+<P><B>Description:</B>
+</P>
+<P>Define a computation that calculates the radial distribution function
+(RDF), also called g(r), and the coordination number for a group of
+particles. Both are calculated in histogram form by binning pairwise
+distances into <I>Nbin</I> bins from 0.0 to the maximum force cutoff
+defined by the <A HREF = "pair_style.html">pair_style</A> command. The bins are of
+uniform size in radial distance. Thus a single bin encompasses a thin
+shell of distances in 3d and a thin ring of distances in 2d.
+</P>
+<P>The <I>itypeN</I> and <I>jtypeN</I> arguments are optional. These arguments
+must come in pairs. If no pairs are listed, then a single histogram
+is computed for g(r) between all atom types. If one or more pairs are
+listed, then a separate histogram is generated for each
+<I>itype</I>,<I>jtype</I> pair.
+</P>
+<P>The <I>itypeN</I> and <I>jtypeN</I> settings can be specified in one of two
+ways. An explicit numeric value can be used, as in the 4th example
+above. Or a wild-card asterisk can be used to specify a range of atom
+types. This takes the form "*" or "*n" or "n*" or "m*n". If N = the
+number of atom types, then an asterisk with no numeric values means
+all types from 1 to N. A leading asterisk means all types from 1 to n
+(inclusive). A trailing asterisk means all types from n to N
+(inclusive). A middle asterisk means all types from m to n
+(inclusive).
+</P>
+<P>If both <I>itypeN</I> and <I>jtypeN</I> are single values, as in the 4th example
+above, this means that a g(r) is computed where atoms of type <I>itypeN</I>
+are the central atom, and atoms of type <I>jtypeN</I> are the distribution
+atom. If either <I>itypeN</I> and <I>jtypeN</I> represent a range of values via
+the wild-card asterisk, as in the 5th example above, this means that a
+g(r) is computed where atoms of any of the range of types represented
+by <I>itypeN</I> are the central atom, and atoms of any of the range of
+types represented by <I>jtypeN</I> are the distribution atom.
+</P>
+<P>Pairwise distances are generated by looping over a pairwise neighbor
+list, just as they would be in a <A HREF = "pair_style.html">pair_style</A>
+computation. The distance between two atoms I and J is included in
+a specific histogram if the following criteria are met:
+</P>
+<UL><LI>atoms I,J are both in the specified fix group
+<LI>the distance between atoms I,J is less than the maximum force cutoff
+<LI>the type of the I atom matches itypeN (one or a range of types)
+<LI>the type of the J atom matches jtypeN (one or a range of types)
+<LI>the I,J interaction is included in the neighbor list
+</UL>
+<P>The last point is relevant for molecular systems with bonds, because
+if two atoms I,J are involved in 1-2, 1-3, 1-4 interactions within the
+molecular topology, their pairwise interaction may be turned off, and
+thus they will not appear in the neighbor list, and will not
+contribute to g(r). More specifically, this is true of pairs with a
+weighting factor of 0.0; pairs with a non-zero weighting factor are
+included. The weighting factors for 1-2, 1-3, and 1-4 pairwise
+interactions is set by the <A HREF = "special_bonds.html">special_bonds</A> command.
+</P>
+<P>It is OK if a particular pairwise distance is included in more than
+one individual histogram, due to the way the <I>itypeN</I> and <I>jtypeN</I>
+arguments are specified.
+</P>
+<P>The g(r) value for a bin is calculated from the histogram count by
+scaling it by the idealized number of how many counts there would be
+if atoms of type <I>jtypeN</I> were uniformly distributed. Thus it
+involves the count of <I>itypeN</I> atoms, the count of <I>jtypeN</I> atoms, the
+volume of the entire simulation box, and the volume of the thin bin
+shell in 3d (or the area of the thin bin ring in 2d).
+</P>
+<P>A coordination number coord(r) is also calculated, which is the sum of
+g(r) values for all bins up to and including the current bin.
+</P>
+<P><B>Output info:</B>
+</P>
+<P>This compute calculates a global array with the number of rows =
+<I>Nbins</I>, and the number of columns = 1 + 2*Npairs, where Npairs is the
+number of I,J pairings specified. The first column has the bin
+coordinate (center of the bin), Each successive set of 2 columns has
+the g(r) and coord(r) values for a specific set of <I>itypeN</I> versus
+<I>jtypeN</I> interactions, as described above. These values can be used
+by any command that uses a global values from a compute as input. See
+<A HREF = "Section_howto.html#4_15">this section</A> for an overview of LAMMPS
+output options.
+</P>
+<P>The array values calculated by this compute are all "intensive", since
+they are normalized, meaning they are independent of the number of
+atoms in the simulation.
+</P>
+<P><B>Restrictions:</B>
+</P>
+<P>The RDF is not computed for distances longer than the force cutoff,
+since processors (in parallel) don't know about atom coordinates for
+atoms further away than that distance. If you want an RDF for larger
+distances, you'll need to post-process a dump file.
+</P>
+<P><B>Related commands:</B>
+</P>
+<P><A HREF = "fix_ave_histo.html">fix ave/histo</A>
+</P>
+<P><B>Default:</B> none
+</P>
+</HTML>
diff --git a/doc/compute_rdf.txt b/doc/compute_rdf.txt
new file mode 100644
index 000000000..9c1205026
--- /dev/null
+++ b/doc/compute_rdf.txt
@@ -0,0 +1,125 @@
+"LAMMPS WWW Site"_lws - "LAMMPS Documentation"_ld - "LAMMPS Commands"_lc :c
+
+:link(lws,http://lammps.sandia.gov)
+:link(ld,Manual.html)
+:link(lc,Section_commands.html#comm)
+
+:line
+
+compute rdf command :h3
+
+[Syntax:]
+
+compute ID group-ID rdf Nbin itype1 jtype1 itype2 jtype2 ... :pre
+
+ID, group-ID are documented in "compute"_compute.html command
+rdf = style name of this compute command
+Nbin = number of RDF bins
+itypeN = central atom type for Nth RDF histogram (see asterisk form below)
+jtypeN = distribution atom type for Nth RDF histogram (see asterisk form below) :ul
+
+[Examples:]
+
+fix 1 all rdf 100
+fix 1 all rdf 100 1 1
+fix 1 all rdf 100 * 3
+fix 1 fluid rdf 500 1 1 1 2 2 1 2 2
+fix 1 fluid rdf 500 1*3 2 5 *10 :pre
+
+[Description:]
+
+Define a computation that calculates the radial distribution function
+(RDF), also called g(r), and the coordination number for a group of
+particles. Both are calculated in histogram form by binning pairwise
+distances into {Nbin} bins from 0.0 to the maximum force cutoff
+defined by the "pair_style"_pair_style.html command. The bins are of
+uniform size in radial distance. Thus a single bin encompasses a thin
+shell of distances in 3d and a thin ring of distances in 2d.
+
+The {itypeN} and {jtypeN} arguments are optional. These arguments
+must come in pairs. If no pairs are listed, then a single histogram
+is computed for g(r) between all atom types. If one or more pairs are
+listed, then a separate histogram is generated for each
+{itype},{jtype} pair.
+
+The {itypeN} and {jtypeN} settings can be specified in one of two
+ways. An explicit numeric value can be used, as in the 4th example
+above. Or a wild-card asterisk can be used to specify a range of atom
+types. This takes the form "*" or "*n" or "n*" or "m*n". If N = the
+number of atom types, then an asterisk with no numeric values means
+all types from 1 to N. A leading asterisk means all types from 1 to n
+(inclusive). A trailing asterisk means all types from n to N
+(inclusive). A middle asterisk means all types from m to n
+(inclusive).
+
+If both {itypeN} and {jtypeN} are single values, as in the 4th example
+above, this means that a g(r) is computed where atoms of type {itypeN}
+are the central atom, and atoms of type {jtypeN} are the distribution
+atom. If either {itypeN} and {jtypeN} represent a range of values via
+the wild-card asterisk, as in the 5th example above, this means that a
+g(r) is computed where atoms of any of the range of types represented
+by {itypeN} are the central atom, and atoms of any of the range of
+types represented by {jtypeN} are the distribution atom.
+
+Pairwise distances are generated by looping over a pairwise neighbor
+list, just as they would be in a "pair_style"_pair_style.html
+computation. The distance between two atoms I and J is included in
+a specific histogram if the following criteria are met:
+
+atoms I,J are both in the specified fix group
+the distance between atoms I,J is less than the maximum force cutoff
+the type of the I atom matches itypeN (one or a range of types)
+the type of the J atom matches jtypeN (one or a range of types)
+the I,J interaction is included in the neighbor list :ul
+
+The last point is relevant for molecular systems with bonds, because
+if two atoms I,J are involved in 1-2, 1-3, 1-4 interactions within the
+molecular topology, their pairwise interaction may be turned off, and
+thus they will not appear in the neighbor list, and will not
+contribute to g(r). More specifically, this is true of pairs with a
+weighting factor of 0.0; pairs with a non-zero weighting factor are
+included. The weighting factors for 1-2, 1-3, and 1-4 pairwise
+interactions is set by the "special_bonds"_special_bonds.html command.
+
+It is OK if a particular pairwise distance is included in more than
+one individual histogram, due to the way the {itypeN} and {jtypeN}
+arguments are specified.
+
+The g(r) value for a bin is calculated from the histogram count by
+scaling it by the idealized number of how many counts there would be
+if atoms of type {jtypeN} were uniformly distributed. Thus it
+involves the count of {itypeN} atoms, the count of {jtypeN} atoms, the
+volume of the entire simulation box, and the volume of the thin bin
+shell in 3d (or the area of the thin bin ring in 2d).
+
+A coordination number coord(r) is also calculated, which is the sum of
+g(r) values for all bins up to and including the current bin.
+
+[Output info:]
+
+This compute calculates a global array with the number of rows =
+{Nbins}, and the number of columns = 1 + 2*Npairs, where Npairs is the
+number of I,J pairings specified. The first column has the bin
+coordinate (center of the bin), Each successive set of 2 columns has
+the g(r) and coord(r) values for a specific set of {itypeN} versus
+{jtypeN} interactions, as described above. These values can be used
+by any command that uses a global values from a compute as input. See
+"this section"_Section_howto.html#4_15 for an overview of LAMMPS
+output options.
+
+The array values calculated by this compute are all "intensive", since
+they are normalized, meaning they are independent of the number of
+atoms in the simulation.
+
+[Restrictions:]
+
+The RDF is not computed for distances longer than the force cutoff,
+since processors (in parallel) don't know about atom coordinates for
+atoms further away than that distance. If you want an RDF for larger
+distances, you'll need to post-process a dump file.
+
+[Related commands:]
+
+"fix ave/histo"_fix_ave_histo.html
+
+[Default:] none