diff --git a/doc/chrony.conf.adoc b/doc/chrony.conf.adoc
index b1e489c..02e3bfa 100644
--- a/doc/chrony.conf.adoc
+++ b/doc/chrony.conf.adoc
@@ -1326,10 +1326,9 @@ It should be noted that this is not the only means of protection against using
 unreliable estimates. At all times, *chronyd* keeps track of both the estimated
 gain or loss rate, and the error bound on the estimate. When a new estimate is
 generated following another measurement from one of the sources, a weighted
-combination algorithm is used to update the master estimate. So if *chronyd*
-has an existing highly-reliable master estimate and a new estimate is generated
-which has large error bounds, the existing master estimate will dominate in the
-new master estimate.
+combination algorithm is used to update the existing estimate. If it has
+significantly smaller error bounds than the new estimate, the existing estimate
+will dominate in the new combined value.
 
 [[maxslewrate]]*maxslewrate* _rate-in-ppm_::
 The *maxslewrate* directive sets the maximum rate at which *chronyd* is allowed
diff --git a/doc/chronyc.adoc b/doc/chronyc.adoc
index 07b4430..46fc46f 100644
--- a/doc/chronyc.adoc
+++ b/doc/chronyc.adoc
@@ -1234,8 +1234,8 @@ deny all
 *local* *off*::
 The *local* command allows *chronyd* to be told that it is to appear as a
 reference source, even if it is not itself properly synchronised to an external
-source. (This can be used on isolated networks, to allow one computer to be a
-master time server with the other computers slaving to it.)
+source. This can be used on isolated networks, to allow a computer to be the
+primary time server for other computers.
 +
 The first form enables the local reference mode on the host. The syntax is
 identical to the <<chrony.conf.adoc#local,*local*>> directive in the
diff --git a/ntp_core.c b/ntp_core.c
index 2bd0104..7462c7b 100644
--- a/ntp_core.c
+++ b/ntp_core.c
@@ -2330,8 +2330,8 @@ NCR_ProcessRxKnown(NCR_Instance inst, NTP_Local_Address *local_addr,
          client mode operation.
 
          This copes with the case for an isolated network where one
-         machine is set by eye and is used as the master, with the
-         other machines pointed at it.  If the master goes down, we
+         machine is set by eye and is used as the primary server, with
+         the other machines pointed at it.  If the server goes down, we
          want to be able to reset its time at startup by relying on
          one of the secondaries to flywheel it. The behaviour coded here
          is required in the secondaries to make this possible. */