diff --git a/CHANGELOG.md b/CHANGELOG.md
index db99d83..b59d84c 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -30,6 +30,8 @@ Even though this refactoring significantly simplified the code, it also introduc
 ### Add
 
 - Add `--all` and `--group` to `pueue log`. [#509](https://github.com/Nukesor/pueue/issues/509)
+- Add `pueue reset [group_names]` to allow resetting individual groups. [#482](https://github.com/Nukesor/pueue/issues/482)
+  This also refactors the way resets are done internally, resulting in a cleaner code architecture.
 
 ## \[3.4.1\] - 2024-06-04
 
diff --git a/pueue/src/client/cli.rs b/pueue/src/client/cli.rs
index 7bbbe71..b52550e 100644
--- a/pueue/src/client/cli.rs
+++ b/pueue/src/client/cli.rs
@@ -455,6 +455,10 @@ https://github.com/Nukesor/pueue/issues/350#issue-1359083118"
 
     /// Kill all tasks, clean up afterwards and reset EVERYTHING!
     Reset {
+        /// If groups are specified, only those specific groups will be reset.
+        #[arg(short, long)]
+        groups: Vec<String>,
+
         /// Don't ask for any confirmation.
         #[arg(short, long)]
         force: bool,
diff --git a/pueue/src/client/client.rs b/pueue/src/client/client.rs
index b62886b..abd8730 100644
--- a/pueue/src/client/client.rs
+++ b/pueue/src/client/client.rs
@@ -545,12 +545,18 @@ impl Client {
                 group: group.clone(),
             }
             .into(),
-            SubCommand::Reset { force, .. } => {
+            SubCommand::Reset { force, groups, .. } => {
                 if self.settings.client.show_confirmation_questions && !force {
                     self.handle_user_confirmation("reset", &Vec::new())?;
                 }
 
-                ResetMessage {}.into()
+                let target = if groups.is_empty() {
+                    ResetTarget::All
+                } else {
+                    ResetTarget::Groups(groups.clone())
+                };
+
+                ResetMessage { target }.into()
             }
             SubCommand::Shutdown => Shutdown::Graceful.into(),
             SubCommand::Parallel {
diff --git a/pueue/src/client/display/group.rs b/pueue/src/client/display/group.rs
index 0da2b6d..3d8ed9e 100644
--- a/pueue/src/client/display/group.rs
+++ b/pueue/src/client/display/group.rs
@@ -57,6 +57,7 @@ pub fn get_group_headline(name: &str, group: &Group, style: &OutputStyle) -> Str
     let status = match group.status {
         GroupStatus::Running => style.style_text("running", Some(Color::Green), None),
         GroupStatus::Paused => style.style_text("paused", Some(Color::Yellow), None),
+        GroupStatus::Reset => style.style_text("resetting", Some(Color::Red), None),
     };
 
     format!("{} ({} parallel): {}", name, group.parallel_tasks, status)
diff --git a/pueue/src/daemon/network/message_handler/mod.rs b/pueue/src/daemon/network/message_handler/mod.rs
index 1b1511e..dac89d2 100644
--- a/pueue/src/daemon/network/message_handler/mod.rs
+++ b/pueue/src/daemon/network/message_handler/mod.rs
@@ -38,7 +38,7 @@ pub fn handle_message(message: Message, state: &SharedState, settings: &Settings
         Message::Parallel(message) => parallel::set_parallel_tasks(message, state),
         Message::Pause(message) => pause::pause(settings, state, message),
         Message::Remove(task_ids) => remove::remove(settings, state, task_ids),
-        Message::Reset(_) => reset::reset(settings, state),
+        Message::Reset(message) => reset::reset(settings, state, message),
         Message::Restart(message) => restart::restart_multiple(settings, state, message),
         Message::Send(message) => send::send(state, message),
         Message::Start(message) => start::start(settings, state, message),
diff --git a/pueue/src/daemon/network/message_handler/reset.rs b/pueue/src/daemon/network/message_handler/reset.rs
index 7adfcf9..e2e9e78 100644
--- a/pueue/src/daemon/network/message_handler/reset.rs
+++ b/pueue/src/daemon/network/message_handler/reset.rs
@@ -1,4 +1,5 @@
-use pueue_lib::state::SharedState;
+use pueue_lib::failure_msg;
+use pueue_lib::state::{GroupStatus, SharedState};
 use pueue_lib::{network::message::*, settings::Settings};
 
 use crate::daemon::process_handler;
@@ -6,9 +7,40 @@ use crate::daemon::process_handler;
 /// Invoked when calling `pueue reset`.
 /// Kill all children by using the `kill` function.
 /// Set the full_reset flag, which will prevent new tasks from being spawned.
-pub fn reset(settings: &Settings, state: &SharedState) -> Message {
+pub fn reset(settings: &Settings, state: &SharedState, message: ResetMessage) -> Message {
     let mut state = state.lock().unwrap();
-    state.full_reset = true;
-    process_handler::kill::kill(settings, &mut state, TaskSelection::All, false, None);
+
+    match message.target {
+        ResetTarget::All => {
+            // Mark all groups to be reset and kill all tasks
+            for (_name, group) in state.groups.iter_mut() {
+                group.status = GroupStatus::Reset;
+            }
+            process_handler::kill::kill(settings, &mut state, TaskSelection::All, false, None);
+        }
+        ResetTarget::Groups(groups) => {
+            // First up, check whether we actually have all requested groups.
+            for name in groups.iter() {
+                let group = state.groups.get(name);
+                if group.is_none() {
+                    return failure_msg!("Group '{name}' doesn't exist.");
+                }
+            }
+
+            // Mark all groups to be reset and kill its tasks
+            for name in groups.iter() {
+                let group = state.groups.get_mut(name).unwrap();
+                group.status = GroupStatus::Reset;
+
+                process_handler::kill::kill(
+                    settings,
+                    &mut state,
+                    TaskSelection::Group(name.to_string()),
+                    false,
+                    None,
+                );
+            }
+        }
+    }
     create_success_message("Everything is being reset right now.")
 }
diff --git a/pueue/src/daemon/process_handler/finish.rs b/pueue/src/daemon/process_handler/finish.rs
index fd08e0e..e0ccee5 100644
--- a/pueue/src/daemon/process_handler/finish.rs
+++ b/pueue/src/daemon/process_handler/finish.rs
@@ -2,6 +2,7 @@ use anyhow::Context;
 use chrono::Local;
 use log::info;
 use pueue_lib::log::clean_log_handles;
+use pueue_lib::state::GroupStatus;
 use pueue_lib::task::{TaskResult, TaskStatus};
 
 use super::*;
@@ -95,8 +96,13 @@ pub fn handle_finished_tasks(settings: &Settings, state: &mut LockedState) {
             pause_on_failure(state, settings, &task.group);
         }
 
-        // Already remove the output files, if the daemon is being reset anyway
-        if state.full_reset {
+        // Already remove the output files, if this group is being reset.
+        if state
+            .groups
+            .get(&task.group)
+            .map(|group| group.status == GroupStatus::Reset)
+            .unwrap_or(true)
+        {
             clean_log_handles(*task_id, &settings.shared.pueue_directory());
         }
     }
diff --git a/pueue/src/daemon/task_handler.rs b/pueue/src/daemon/task_handler.rs
index 90d2109..0ce7119 100644
--- a/pueue/src/daemon/task_handler.rs
+++ b/pueue/src/daemon/task_handler.rs
@@ -6,7 +6,6 @@ use chrono::prelude::*;
 use log::{error, info};
 
 use pueue_lib::children::Children;
-use pueue_lib::log::*;
 use pueue_lib::network::message::*;
 use pueue_lib::network::protocol::socket_cleanup;
 use pueue_lib::settings::Settings;
@@ -14,7 +13,7 @@ use pueue_lib::state::{Group, GroupStatus, SharedState};
 use pueue_lib::task::{TaskResult, TaskStatus};
 
 use crate::daemon::pid::cleanup_pid_file;
-use crate::daemon::state_helper::{reset_state, save_state};
+use crate::daemon::state_helper::save_state;
 use crate::ok_or_shutdown;
 
 use super::callbacks::{check_callbacks, spawn_callback};
@@ -46,26 +45,24 @@ pub async fn run(state: SharedState, settings: Settings) -> Result<()> {
     }
 
     loop {
-        {
+        'mutex_block: {
             let mut state = state.lock().unwrap();
 
             check_callbacks(&mut state);
             handle_finished_tasks(&settings, &mut state);
+
+            // Check if we're in shutdown.
+            // If all tasks are killed, we do some cleanup and exit.
+            if state.shutdown.is_some() {
+                handle_shutdown(&settings, &mut state);
+                break 'mutex_block;
+            }
+
+            // If we aren't in shutdown mode, do the usual stuff
+            handle_group_resets(&settings, &mut state);
             enqueue_delayed_tasks(&settings, &mut state);
             check_failed_dependencies(&settings, &mut state);
-
-            if state.shutdown.is_some() {
-                // Check if we're in shutdown.
-                // If all tasks are killed, we do some cleanup and exit.
-                handle_shutdown(&settings, &mut state);
-            } else if state.full_reset {
-                // Wait until all tasks are killed.
-                // Once they are, reset everything and go back to normal
-                handle_reset(&settings, &mut state);
-            } else {
-                // Only start new tasks, if we aren't in the middle of a reset or shutdown.
-                spawn_new(&settings, &mut state);
-            }
+            spawn_new(&settings, &mut state);
         }
 
         tokio::time::sleep(Duration::from_millis(300)).await;
@@ -106,20 +103,28 @@ fn handle_shutdown(settings: &Settings, state: &mut LockedState) {
 /// and no new tasks will be spawned.
 /// This function checks, if all killed children have been handled.
 /// If that's the case, completely reset the state
-fn handle_reset(settings: &Settings, state: &mut LockedState) {
-    // Don't do any reset logic, if we aren't in reset mode or if some children are still up.
-    if state.children.has_active_tasks() {
-        return;
+fn handle_group_resets(_settings: &Settings, state: &mut LockedState) {
+    let groups_to_reset: Vec<String> = state
+        .groups
+        .iter()
+        .filter(|(_name, group)| group.status == GroupStatus::Reset)
+        .map(|(name, _)| name.to_string())
+        .collect();
+
+    for name in groups_to_reset.iter() {
+        // Don't do any reset logic, if there're still some children are still up.
+        if state.children.has_group_active_tasks(name) {
+            continue;
+        }
+
+        // Remove all tasks that belong to the group to reset
+        state.tasks.retain(|_id, task| &task.group != name);
+
+        // Restart the group, now that it's devoid of tasks.
+        if let Some(group) = state.groups.get_mut(name) {
+            group.status = GroupStatus::Running;
+        }
     }
-
-    if let Err(error) = reset_state(state, settings) {
-        error!("Failed to reset state with error: {error:?}");
-    };
-
-    if let Err(error) = reset_task_log_directory(&settings.shared.pueue_directory()) {
-        panic!("Error while resetting task log directory: {error}");
-    };
-    state.full_reset = false;
 }
 
 /// As time passes, some delayed tasks may need to be enqueued.
diff --git a/pueue/tests/daemon/integration/reset.rs b/pueue/tests/daemon/integration/reset.rs
index ac4fd84..9e5bfa9 100644
--- a/pueue/tests/daemon/integration/reset.rs
+++ b/pueue/tests/daemon/integration/reset.rs
@@ -1,9 +1,9 @@
 use anyhow::{Context, Result};
-use pueue_lib::network::message::*;
+use pueue_lib::{network::message::*, state::GroupStatus};
 
 use crate::helper::*;
 
-/// A reset command kills all tasks and forces a clean state.
+/// A reset command kills all tasks and forces a clean state accross groups.
 #[tokio::test(flavor = "multi_thread", worker_threads = 2)]
 async fn test_reset() -> Result<()> {
     let daemon = daemon().await?;
@@ -12,21 +12,128 @@ async fn test_reset() -> Result<()> {
     // Start a long running task and make sure it's started
     add_task(shared, "ls").await?;
     add_task(shared, "failed").await?;
-    add_task(shared, "sleep 60").await?;
+    add_task_to_group(shared, "sleep 60", "test_2").await?;
     add_task(shared, "ls").await?;
     wait_for_task_condition(shared, 2, |task| task.is_running()).await?;
 
-    // Reset the daemon
-    send_message(shared, ResetMessage {})
-        .await
-        .context("Failed to send Start tasks message")?;
+    // Reset all groups of the daemon
+    send_message(
+        shared,
+        ResetMessage {
+            target: ResetTarget::All,
+        },
+    )
+    .await
+    .context("Failed to send Start tasks message")?;
 
-    // Resetting is asynchronous, wait for the first task to disappear.
+    // Resetting is asynchronous, wait for all task to disappear.
     wait_for_task_absence(shared, 0).await?;
+    wait_for_task_absence(shared, 1).await?;
+    wait_for_task_absence(shared, 2).await?;
+    wait_for_task_absence(shared, 3).await?;
 
     // All tasks should have been removed.
     let state = get_state(shared).await?;
     assert!(state.tasks.is_empty(),);
 
+    // Both groups should be running.
+    assert_eq!(
+        state.groups.get("default").unwrap().status,
+        GroupStatus::Running
+    );
+    assert_eq!(
+        state.groups.get("test_2").unwrap().status,
+        GroupStatus::Running
+    );
+
+    Ok(())
+}
+
+/// A reset command kills all tasks and forces a clean state.
+#[tokio::test(flavor = "multi_thread", worker_threads = 2)]
+async fn test_reset_single_group() -> Result<()> {
+    let daemon = daemon().await?;
+    let shared = &daemon.settings.shared;
+
+    // Start a long running task and make sure it's started
+    add_task(shared, "ls").await?;
+    add_task(shared, "failed").await?;
+    add_task_to_group(shared, "sleep 60", "test_2").await?;
+    add_task_to_group(shared, "sleep 60", "test_3").await?;
+    wait_for_task_condition(shared, 2, |task| task.is_running()).await?;
+
+    // Reset only the test_2 of the daemon.
+    send_message(
+        shared,
+        ResetMessage {
+            target: ResetTarget::Groups(vec!["test_2".to_string()]),
+        },
+    )
+    .await
+    .context("Failed to send Start tasks message")?;
+
+    // Resetting is asynchronous, wait for the third task to disappear.
+    wait_for_task_absence(shared, 2).await?;
+
+    // All tasks should have been removed.
+    let state = get_state(shared).await?;
+    assert_eq!(
+        state.tasks.len(),
+        3,
+        "Only a single task should have been removed"
+    );
+
+    assert_eq!(
+        state.groups.get("test_2").unwrap().status,
+        GroupStatus::Running
+    );
+
+    Ok(())
+}
+
+/// A reset command kills all tasks and forces a clean state.
+#[tokio::test(flavor = "multi_thread", worker_threads = 2)]
+async fn test_reset_multiple_groups() -> Result<()> {
+    let daemon = daemon().await?;
+    let shared = &daemon.settings.shared;
+
+    // Start a long running task and make sure it's started
+    add_task(shared, "ls").await?;
+    add_task(shared, "failed").await?;
+    add_task_to_group(shared, "sleep 60", "test_2").await?;
+    add_task_to_group(shared, "sleep 60", "test_3").await?;
+    wait_for_task_condition(shared, 2, |task| task.is_running()).await?;
+
+    // Reset only the test_2 of the daemon.
+    send_message(
+        shared,
+        ResetMessage {
+            target: ResetTarget::Groups(vec!["test_2".to_string(), "test_3".to_string()]),
+        },
+    )
+    .await
+    .context("Failed to send Start tasks message")?;
+
+    // Resetting is asynchronous, wait for the third task to disappear.
+    wait_for_task_absence(shared, 2).await?;
+    wait_for_task_absence(shared, 3).await?;
+
+    // All tasks should have been removed.
+    let state = get_state(shared).await?;
+    assert_eq!(
+        state.tasks.len(),
+        2,
+        "Only a two task should have been removed"
+    );
+
+    assert_eq!(
+        state.groups.get("test_2").unwrap().status,
+        GroupStatus::Running
+    );
+    assert_eq!(
+        state.groups.get("test_3").unwrap().status,
+        GroupStatus::Running
+    );
+
     Ok(())
 }
diff --git a/pueue_lib/src/children.rs b/pueue_lib/src/children.rs
index f16f18d..5055c2d 100644
--- a/pueue_lib/src/children.rs
+++ b/pueue_lib/src/children.rs
@@ -15,6 +15,16 @@ impl Children {
         self.0.iter().any(|(_, pool)| !pool.is_empty())
     }
 
+    /// Returns whether there are any active tasks for the given group.
+    ///
+    /// Returns `false` if the group cannot be found.
+    pub fn has_group_active_tasks(&self, group: &str) -> bool {
+        self.0
+            .get(group)
+            .map(|pool| !pool.is_empty())
+            .unwrap_or(false)
+    }
+
     /// A convenience function to check whether there's child with a given task_id.
     /// We have to do a nested linear search, as these datastructure aren't indexed via task_ids.
     pub fn has_child(&self, task_id: usize) -> bool {
diff --git a/pueue_lib/src/log.rs b/pueue_lib/src/log.rs
index b119be7..aed9cf6 100644
--- a/pueue_lib/src/log.rs
+++ b/pueue_lib/src/log.rs
@@ -1,4 +1,4 @@
-use std::fs::{read_dir, remove_file, File};
+use std::fs::{remove_file, File};
 use std::io::{self, prelude::*, Read, SeekFrom};
 use std::path::{Path, PathBuf};
 
@@ -104,22 +104,6 @@ pub fn read_last_log_file_lines(
     Ok(read_last_lines(&mut file, lines))
 }
 
-/// Remove all files in the log directory.
-pub fn reset_task_log_directory(pueue_dir: &Path) -> Result<(), Error> {
-    let task_log_dir = pueue_dir.join("task_logs");
-
-    let files = read_dir(&task_log_dir)
-        .map_err(|err| Error::IoPathError(task_log_dir, "reading task log files", err))?;
-
-    for file in files.flatten() {
-        if let Err(err) = remove_file(file.path()) {
-            error!("Failed to delete log file: {err}");
-        }
-    }
-
-    Ok(())
-}
-
 /// Read the last `amount` lines of a file to a string.
 ///
 /// Only use this for logic that doesn't stream from daemon to client!
diff --git a/pueue_lib/src/network/message.rs b/pueue_lib/src/network/message.rs
index 507b490..9e49779 100644
--- a/pueue_lib/src/network/message.rs
+++ b/pueue_lib/src/network/message.rs
@@ -280,7 +280,17 @@ pub struct GroupResponseMessage {
 impl_into_message!(GroupResponseMessage, Message::GroupResponse);
 
 #[derive(PartialEq, Eq, Clone, Debug, Deserialize, Serialize)]
-pub struct ResetMessage {}
+pub enum ResetTarget {
+    // Reset all groups
+    All,
+    // Reset a list of specific groups
+    Groups(Vec<String>),
+}
+
+#[derive(PartialEq, Eq, Clone, Debug, Deserialize, Serialize)]
+pub struct ResetMessage {
+    pub target: ResetTarget,
+}
 
 impl_into_message!(ResetMessage, Message::Reset);
 
diff --git a/pueue_lib/src/state.rs b/pueue_lib/src/state.rs
index a5eb65c..1be1b03 100644
--- a/pueue_lib/src/state.rs
+++ b/pueue_lib/src/state.rs
@@ -19,6 +19,9 @@ pub type SharedState = Arc<Mutex<State>>;
 pub enum GroupStatus {
     Running,
     Paused,
+    // This state is set, if this group is being reset.
+    // This means that all tasks are being killed and removed.
+    Reset,
 }
 
 /// The representation of a group.
@@ -62,11 +65,7 @@ pub struct State {
     /// This is runtime state and won't be serialised to disk.
     #[serde(default, skip)]
     pub shutdown: Option<Shutdown>,
-    /// A simple flag which is used to signal that we're currently doing a full reset of the daemon.
-    /// This flag prevents new tasks from being spawned.
-    /// This is runtime state and won't be serialised to disk.
-    #[serde(default, skip)]
-    pub full_reset: bool,
+
     /// Pueue's subprocess and worker pool representation.
     /// Take a look at [Children] for more info.
     /// This is runtime state and won't be serialised to disk.
@@ -84,7 +83,6 @@ impl Clone for State {
             tasks: self.tasks.clone(),
             groups: self.groups.clone(),
             shutdown: self.shutdown.clone(),
-            full_reset: self.full_reset,
             ..Default::default()
         }
     }
@@ -94,10 +92,7 @@ impl Clone for State {
 impl Eq for State {}
 impl PartialEq for State {
     fn eq(&self, other: &Self) -> bool {
-        self.tasks == other.tasks
-            && self.groups == other.groups
-            && self.shutdown == other.shutdown
-            && self.full_reset == other.full_reset
+        self.tasks == other.tasks && self.groups == other.groups && self.shutdown == other.shutdown
     }
 }