diff --git a/API/Schema/Jobs/DownloadNewChaptersJob.cs b/API/Schema/Jobs/DownloadNewChaptersJob.cs
index 6cc82b8..b00472e 100644
--- a/API/Schema/Jobs/DownloadNewChaptersJob.cs
+++ b/API/Schema/Jobs/DownloadNewChaptersJob.cs
@@ -12,11 +12,22 @@ public class DownloadNewChaptersJob(ulong recurrenceMs, string mangaId, string?
     
     protected override IEnumerable<Job> RunInternal(PgsqlContext context)
     {
-        Manga m = Manga ?? context.Manga.Find(MangaId)!;
-        MangaConnector connector = m.MangaConnector ?? context.MangaConnectors.Find(m.MangaConnectorId)!;
-        Chapter[] newChapters = connector.GetNewChapters(m);
+        /*
+         * For some reason, directly using Manga from above instead of finding it again causes DBContext to consider
+         * Manga as a new entity and Postgres throws a Duplicate PK exception.
+         * m.MangaConnector does not have this issue (IDK why).
+         */ 
+        Manga m = context.Manga.Find(MangaId)!; 
+        MangaConnector connector = context.MangaConnectors.Find(m.MangaConnectorId)!;
+        // This gets all chapters that are not downloaded
+        Chapter[] allNewChapters = connector.GetNewChapters(m);
+        
+        // This filters out chapters that are not downloaded but already exist in the DB
+        string[] chapterIds = context.Chapters.Where(chapter => chapter.ParentMangaId == m.MangaId).Select(chapter => chapter.ChapterId).ToArray();
+        Chapter[] newChapters = allNewChapters.Where(chapter => !chapterIds.Contains(chapter.ChapterId)).ToArray();
         context.Chapters.AddRangeAsync(newChapters).Wait();
         context.SaveChangesAsync().Wait();
-        return newChapters.Select(chapter => new DownloadSingleChapterJob(chapter.ChapterId, this.JobId));
+        
+        return allNewChapters.Select(chapter => new DownloadSingleChapterJob(chapter.ChapterId, this.JobId));
     }
 }
\ No newline at end of file