diff --git a/plat/imx/common/include/sci/sci_rpc.h b/plat/imx/common/include/sci/sci_rpc.h
index 60dbc27..b6adf33 100644
--- a/plat/imx/common/include/sci/sci_rpc.h
+++ b/plat/imx/common/include/sci/sci_rpc.h
@@ -100,7 +100,7 @@
 void sc_call_rpc(sc_ipc_t ipc, sc_rpc_msg_t *msg, bool no_resp);
 
 /*!
- * This is an internal function to dispath an RPC call that has
+ * This is an internal function to dispatch an RPC call that has
  * arrived via IPC over an MU. It is called by server-side SCFW.
  *
  * @param[in]     mu          MU message arrived on
diff --git a/plat/imx/common/include/sci/svc/pad/sci_pad_api.h b/plat/imx/common/include/sci/svc/pad/sci_pad_api.h
index dc23eed..ac93aae 100644
--- a/plat/imx/common/include/sci/svc/pad/sci_pad_api.h
+++ b/plat/imx/common/include/sci/svc/pad/sci_pad_api.h
@@ -42,7 +42,7 @@
  *
  * Pads are managed as a resource by the Resource Manager (RM).  They have
  * assigned owners and only the owners can configure the pads. Some of the
- * pads are reserved for use by the SCFW itself and this can be overriden
+ * pads are reserved for use by the SCFW itself and this can be overridden
  * with the implementation of board_config_sc(). Additionally, pads may
  * be assigned to various other partitions via the implementation of
  * board_system_config().
@@ -156,7 +156,7 @@
  * This type is used to declare a pad low-power isolation config.
  * ISO_LATE is the most common setting. ISO_EARLY is only used when
  * an output pad is directly determined by another input pad. The
- * other two are only used when SW wants to directly contol isolation.
+ * other two are only used when SW wants to directly control isolation.
  */
 typedef uint8_t sc_pad_iso_t;
 
diff --git a/plat/imx/common/include/sci/svc/pm/sci_pm_api.h b/plat/imx/common/include/sci/svc/pm/sci_pm_api.h
index 76ca5c4..1364795 100644
--- a/plat/imx/common/include/sci/svc/pm/sci_pm_api.h
+++ b/plat/imx/common/include/sci/svc/pm/sci_pm_api.h
@@ -294,7 +294,7 @@
  * Note some resources are still not accessible even when powered up if bus
  * transactions go through a fabric not powered up. Examples of this are
  * resources in display and capture subsystems which require the display
- * controller or the imaging subsytem to be powered up first.
+ * controller or the imaging subsystem to be powered up first.
  *
  * Not that resources are grouped into power domains by the underlying
  * hardware. If any resource in the domain is on, the entire power domain