diff --git a/src/libcollections/bit.rs b/src/libcollections/bit.rs
index f44f700ae5e42..c0e37f7157a20 100644
--- a/src/libcollections/bit.rs
+++ b/src/libcollections/bit.rs
@@ -283,7 +283,7 @@ impl Bitv {
         x != 0
     }
 
-    /// Sets the value of a bit at a index `i`.
+    /// Sets the value of a bit at an index `i`.
     ///
     /// # Panics
     ///
@@ -582,7 +582,7 @@ impl Bitv {
     ///
     /// # Panics
     ///
-    /// Panics if the the `Bitv` and slice are of different length.
+    /// Panics if the `Bitv` and slice are of different length.
     ///
     /// # Example
     ///
diff --git a/src/libcollections/btree/map.rs b/src/libcollections/btree/map.rs
index 57ca585f15e7b..93bcf49961b62 100644
--- a/src/libcollections/btree/map.rs
+++ b/src/libcollections/btree/map.rs
@@ -53,7 +53,7 @@ use ring_buf::RingBuf;
 /// and possibly other factors. Using linear search, searching for a random element is expected
 /// to take O(B log<sub>B</sub>n) comparisons, which is generally worse than a BST. In practice,
 /// however, performance is excellent. `BTreeMap` is able to readily outperform `TreeMap` under
-/// many workloads, and is competetive where it doesn't. BTreeMap also generally *scales* better
+/// many workloads, and is competitive where it doesn't. BTreeMap also generally *scales* better
 /// than TreeMap, making it more appropriate for large datasets.
 ///
 /// However, `TreeMap` may still be more appropriate to use in many contexts. If elements are very
diff --git a/src/libcollections/btree/node.rs b/src/libcollections/btree/node.rs
index a15d31ba2842a..378657b93cb89 100644
--- a/src/libcollections/btree/node.rs
+++ b/src/libcollections/btree/node.rs
@@ -43,7 +43,7 @@ pub struct Node<K, V> {
     // theory, if we take full control of allocation like HashMap's RawTable does,
     // and restrict leaves to max size 256 (not unreasonable for a btree node) we can cut
     // this down to just (ptr, cap: u8, size: u8, is_leaf: bool). With generic
-    // integer arguments, cap can even move into the the type, reducing this just to
+    // integer arguments, cap can even move into the type, reducing this just to
     // (ptr, size, is_leaf). This could also have cache benefits for very small nodes, as keys
     // could bleed into edges and vals.
     //
diff --git a/src/libcollections/string.rs b/src/libcollections/string.rs
index 3c75198a368bf..5866c067b2dfe 100644
--- a/src/libcollections/string.rs
+++ b/src/libcollections/string.rs
@@ -549,7 +549,7 @@ impl String {
     ///
     /// # Warning
     ///
-    /// This is a O(n) operation as it requires copying every element in the
+    /// This is an O(n) operation as it requires copying every element in the
     /// buffer.
     ///
     /// # Panics
@@ -586,7 +586,7 @@ impl String {
     ///
     /// # Warning
     ///
-    /// This is a O(n) operation as it requires copying every element in the
+    /// This is an O(n) operation as it requires copying every element in the
     /// buffer.
     ///
     /// # Panics
diff --git a/src/libcollections/trie/map.rs b/src/libcollections/trie/map.rs
index da53b5e9166ab..93ed0b0461634 100644
--- a/src/libcollections/trie/map.rs
+++ b/src/libcollections/trie/map.rs
@@ -1056,7 +1056,7 @@ impl<'a, T> VacantEntry<'a, T> {
             search_stack.map.root.count = temp;
             value_ref
         }
-        // Otherwise, find the predeccessor of the last stack node, and insert as normal.
+        // Otherwise, find the predecessor of the last stack node, and insert as normal.
         else {
             match *search_stack.get_ref(old_length - 2) {
                 Internal(box ref mut parent) => {
@@ -1739,7 +1739,7 @@ mod test {
                     // Update it to i^3 using the returned mutable reference.
                     *inserted_val = i * i * i;
                 },
-                _ => panic!("Non-existant key found.")
+                _ => panic!("Non-existent key found.")
             }
             assert_eq!(map.get(&i).unwrap(), &(i * i * i));
         }