diff --git a/c/0128-longest-consecutive-sequence.c b/c/0128-longest-consecutive-sequence.c
new file mode 100644
index 000000000..a750ba4b5
--- /dev/null
+++ b/c/0128-longest-consecutive-sequence.c
@@ -0,0 +1,158 @@
+#define INIT_HASH_SIZE 4096
+
+// Represent an element in a hash table.
+typedef struct Hash {
+    int key;
+    struct Hash *next;
+} Hash;
+
+Hash **InitHash() {
+    Hash **h = (Hash**)calloc(INIT_HASH_SIZE, sizeof(Hash*));
+    assert(h);
+    return h;
+}
+
+Hash *NewHash(int key) {
+    Hash *h = (Hash*)malloc(sizeof(Hash));
+    assert(h);
+    h->key = key;
+    h->next = NULL;
+    return h;
+}
+
+int HashKey(int key) {
+    while(key < 0) key += INIT_HASH_SIZE;
+    return (key % INIT_HASH_SIZE);
+}
+
+void AddHash(Hash **hash, int key) {
+    assert(hash);
+    int hash_key = HashKey(key);
+    Hash **head = hash + hash_key;
+    if (!*head) {
+        *head = NewHash(key);
+    } else {
+        Hash *h = *head;
+        while (h) {
+            if (h->key == key) {
+                return;
+            } else if (!h->next) {
+                h->next = NewHash(key);
+                return;
+            }
+            h = h->next;
+        }
+    }
+}
+
+// If hash not round return 0 on purpose.
+int GetHash(Hash **hash, int key) {
+    assert(hash);
+    Hash * h = hash[HashKey(key)];
+    while(h) {
+        if (h->key == key) {
+            return 1;
+        }
+        h = h->next;
+    }
+
+    return 0;
+}
+
+int longestConsecutive(int* nums, int numsSize){
+    Hash **hash = InitHash();
+    int i, len, n, maxLen = 0;
+
+    for (i = 0; i < numsSize; i++) {
+        AddHash(hash, nums[i]);
+    }
+    for (i = 0; i < numsSize; i++) {
+        n = nums[i];
+        len = 1;
+        if (GetHash(hash, n - 1) == 0) {
+            while(GetHash(hash, ++n) == 1) {
+                len++;
+            }
+        }
+        maxLen = (len > maxLen) ? len : maxLen;
+    }
+
+    return maxLen;
+}
+
+//
+// According to the official solution, using hash map will be considered as O(n).
+//
+// Alternative solution 1 - With uthash, the runtime is poorer. It is a little surprise.
+//
+// typedef struct Hash {
+//     int num;
+//     UT_hash_handle hh;
+// } Hash;
+//
+// void AddHash(Hash **hash, int num) {
+//     Hash *entry;
+//     HASH_FIND_INT(*hash, &num, entry);
+//     if (!entry) {
+//         entry = malloc(sizeof(Hash));
+//         entry->num = num;
+//         HASH_ADD_INT(*hash, num, entry);
+//     }
+// }
+//
+// Hash * GetHash(Hash **hash, int num) {
+//     Hash *entry = NULL;
+//     HASH_FIND_INT(*hash, &num, entry);
+//     return entry;
+// }
+//
+// int longestConsecutive(int* nums, int numsSize){
+//     Hash *root = NULL;
+//     int i, len, n, maxLen = 0;
+//
+//     for (i = 0; i < numsSize; i++) {
+//         AddHash(&root, nums[i]);
+//     }
+//     for (i = 0; i < numsSize; i++) {
+//         n = nums[i];
+//         len = 1;
+//         if (!GetHash(&root, n - 1)) {
+//             while(GetHash(&root, ++n)) {
+//                 len++;
+//             }
+//         }
+//         maxLen = (len > maxLen) ? len : maxLen;
+//     }
+//
+//     return maxLen;
+// }
+//
+//
+// Alternative solution 2 - With sorting, O(nlogn), the runtime is the best.
+//
+// int compare(const void *a, const void *b) {
+//     return *(int*)a - *(int*)b;
+// }
+//
+// int longestConsecutive(int* nums, int numsSize){
+//     int i, len, max;
+//     if (numsSize <= 0) {
+//         return 0;
+//     }
+//     len = 1;
+//     max = 1;
+//     qsort(nums, numsSize, sizeof(int), compare);
+//     for (i = 0; i < numsSize - 1; i++) {
+//         if (nums[i] + 1 == nums[i + 1]) {
+//             len++;
+//         } else if (nums[i] == nums[i + 1]){
+//             continue;
+//         } else {
+//             len = 1;
+//         }
+//         max = (max > len) ? max : len;
+//     }
+//
+//     return max;
+// }
+//