diff --git a/CHANGELOG.md b/CHANGELOG.md
index 4b27c9e4..e6b4afc5 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -14,8 +14,9 @@ The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.0.0/).
 * Turn off `Window` title bars with `window.title_bar(false)`.
 * `ImageButton` - `ui.add(ImageButton::new(...))`.
 * `ui.vertical_centered` and `ui.vertical_centered_justified`.
-* Mouse-over explanation to duplicate ID warning
+* Mouse-over explanation to duplicate ID warning.
 * You can now easily constrain Egui to a portion of the screen using `RawInput::screen_rect`.
+* You can now control the minimum and maixumum number of decimals to show in a `Slider` or `DragValue`.
 
 ### Changed 🔧
 
@@ -23,6 +24,7 @@ The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.0.0/).
 * `SidePanel::left` and `TopPanel::top` now takes `impl Hash` as first argument.
 * `ui.image` now takes `impl Into<Vec2>` as a `size` argument.
 * Made some more fields of `RawInput` optional.
+* `Slider` and `DragValue` uses fewer decimals by default. See the full precision by hovering over the value.
 
 ### Deprecated
 * Deprecated `RawInput::screen_size` - use `RawInput::screen_rect` instead.
diff --git a/egui/src/demos/sliders.rs b/egui/src/demos/sliders.rs
index 7cf22297..20efe6aa 100644
--- a/egui/src/demos/sliders.rs
+++ b/egui/src/demos/sliders.rs
@@ -67,7 +67,10 @@ impl Sliders {
                     .text("f64 demo slider"),
             );
 
-            ui.label("Sliders will automatically figure out how many decimals to show.");
+            ui.label(
+                "Sliders will intelligently pick how many decimals to show. \
+                You can always see the full precision value by hovering the value.",
+            );
 
             if ui.button("Assign PI").clicked {
                 self.value = std::f64::consts::PI;
diff --git a/egui/src/introspection.rs b/egui/src/introspection.rs
index 9d967248..f22b0ff2 100644
--- a/egui/src/introspection.rs
+++ b/egui/src/introspection.rs
@@ -60,7 +60,7 @@ impl paint::FontDefinitions {
             // TODO: radio button for family
             ui.add(
                 Slider::f32(size, 4.0..=40.0)
-                    .precision(0)
+                    .max_decimals(0)
                     .text(format!("{:?}", text_style)),
             );
         }
diff --git a/egui/src/math/mod.rs b/egui/src/math/mod.rs
index 2cb8f867..f43ca576 100644
--- a/egui/src/math/mod.rs
+++ b/egui/src/math/mod.rs
@@ -121,26 +121,41 @@ pub fn ease_in_ease_out(t: f32) -> f32 {
 pub const TAU: f32 = 2.0 * std::f32::consts::PI;
 
 /// Round a value to the given number of decimal places.
-pub fn round_to_precision(value: f64, decimal_places: usize) -> f64 {
+pub fn round_to_decimals(value: f64, decimal_places: usize) -> f64 {
     // This is a stupid way of doing this, but stupid works.
     format!("{:.*}", decimal_places, value)
         .parse()
         .unwrap_or(value)
 }
 
-pub fn format_with_minimum_precision(value: f32, precision: usize) -> String {
-    debug_assert!(precision < 100);
-    let precision = precision.min(16);
-    let text = format!("{:.*}", precision, value);
-    let epsilon = 16.0 * f32::EPSILON; // margin large enough to handle most peoples round-tripping needs
-    if almost_equal(text.parse::<f32>().unwrap(), value, epsilon) {
-        // Enough precision to show the value accurately - good!
-        text
+pub fn format_with_minimum_decimals(value: f64, decimals: usize) -> String {
+    format_with_decimals_in_range(value, decimals..=6)
+}
+
+pub fn format_with_decimals_in_range(value: f64, decimal_range: RangeInclusive<usize>) -> String {
+    let min_decimals = *decimal_range.start();
+    let max_decimals = *decimal_range.end();
+    debug_assert!(min_decimals <= max_decimals);
+    debug_assert!(max_decimals < 100);
+    let max_decimals = max_decimals.min(16);
+    let min_decimals = min_decimals.min(max_decimals);
+
+    if min_decimals == max_decimals {
+        format!("{:.*}", max_decimals, value)
     } else {
+        // Ugly/slow way of doing this. TODO: clean up precision.
+        for decimals in min_decimals..max_decimals {
+            let text = format!("{:.*}", decimals, value);
+            let epsilon = 16.0 * f32::EPSILON; // margin large enough to handle most peoples round-tripping needs
+            if almost_equal(text.parse::<f32>().unwrap(), value as f32, epsilon) {
+                // Enough precision to show the value accurately - good!
+                return text;
+            }
+        }
         // The value has more precision than we expected.
         // Probably the value was set not by the slider, but from outside.
         // In any case: show the full value
-        value.to_string()
+        format!("{:.*}", max_decimals, value)
     }
 }
 
@@ -162,12 +177,13 @@ pub fn almost_equal(a: f32, b: f32, epsilon: f32) -> bool {
 #[allow(clippy::approx_constant)]
 #[test]
 fn test_format() {
-    assert_eq!(format_with_minimum_precision(1_234_567.0, 0), "1234567");
-    assert_eq!(format_with_minimum_precision(1_234_567.0, 1), "1234567.0");
-    assert_eq!(format_with_minimum_precision(3.14, 2), "3.14");
+    assert_eq!(format_with_minimum_decimals(1_234_567.0, 0), "1234567");
+    assert_eq!(format_with_minimum_decimals(1_234_567.0, 1), "1234567.0");
+    assert_eq!(format_with_minimum_decimals(3.14, 2), "3.14");
+    assert_eq!(format_with_minimum_decimals(3.14, 3), "3.140");
     assert_eq!(
-        format_with_minimum_precision(std::f32::consts::PI, 2),
-        "3.1415927"
+        format_with_minimum_decimals(std::f64::consts::PI, 2),
+        "3.14159"
     );
 }
 
diff --git a/egui/src/widgets/drag_value.rs b/egui/src/widgets/drag_value.rs
index 8bf636df..b00438c4 100644
--- a/egui/src/widgets/drag_value.rs
+++ b/egui/src/widgets/drag_value.rs
@@ -23,6 +23,8 @@ pub struct DragValue<'a> {
     prefix: String,
     suffix: String,
     range: RangeInclusive<f64>,
+    min_decimals: usize,
+    max_decimals: Option<usize>,
 }
 
 impl<'a> DragValue<'a> {
@@ -33,51 +35,47 @@ impl<'a> DragValue<'a> {
             prefix: Default::default(),
             suffix: Default::default(),
             range: f64::NEG_INFINITY..=f64::INFINITY,
+            min_decimals: 0,
+            max_decimals: None,
         }
     }
 
     pub fn f32(value: &'a mut f32) -> Self {
-        Self {
-            ..Self::from_get_set(move |v: Option<f64>| {
-                if let Some(v) = v {
-                    *value = v as f32
-                }
-                *value as f64
-            })
-        }
+        Self::from_get_set(move |v: Option<f64>| {
+            if let Some(v) = v {
+                *value = v as f32
+            }
+            *value as f64
+        })
     }
 
     pub fn f64(value: &'a mut f64) -> Self {
-        Self {
-            ..Self::from_get_set(move |v: Option<f64>| {
-                if let Some(v) = v {
-                    *value = v
-                }
-                *value
-            })
-        }
+        Self::from_get_set(move |v: Option<f64>| {
+            if let Some(v) = v {
+                *value = v
+            }
+            *value
+        })
     }
 
     pub fn u8(value: &'a mut u8) -> Self {
-        Self {
-            ..Self::from_get_set(move |v: Option<f64>| {
-                if let Some(v) = v {
-                    *value = v.round() as u8;
-                }
-                *value as f64
-            })
-        }
+        Self::from_get_set(move |v: Option<f64>| {
+            if let Some(v) = v {
+                *value = v.round() as u8;
+            }
+            *value as f64
+        })
+        .max_decimals(0)
     }
 
     pub fn i32(value: &'a mut i32) -> Self {
-        Self {
-            ..Self::from_get_set(move |v: Option<f64>| {
-                if let Some(v) = v {
-                    *value = v.round() as i32;
-                }
-                *value as f64
-            })
-        }
+        Self::from_get_set(move |v: Option<f64>| {
+            if let Some(v) = v {
+                *value = v.round() as i32;
+            }
+            *value as f64
+        })
+        .max_decimals(0)
     }
 
     /// How much the value changes when dragged one point (logical pixel).
@@ -103,6 +101,35 @@ impl<'a> DragValue<'a> {
         self.suffix = suffix.to_string();
         self
     }
+
+    // TODO: we should also have a "min precision".
+    /// Set a minimum number of decimals to display.
+    /// Normally you don't need to pick a precision, as the slider will intelligently pick a precision for you.
+    /// Regardless of precision the slider will use "smart aim" to help the user select nice, round values.
+    pub fn min_decimals(mut self, min_decimals: usize) -> Self {
+        self.min_decimals = min_decimals;
+        self
+    }
+
+    // TODO: we should also have a "max precision".
+    /// Set a maximum number of decimals to display.
+    /// Values will also be rounded to this number of decimals.
+    /// Normally you don't need to pick a precision, as the slider will intelligently pick a precision for you.
+    /// Regardless of precision the slider will use "smart aim" to help the user select nice, round values.
+    pub fn max_decimals(mut self, max_decimals: usize) -> Self {
+        self.max_decimals = Some(max_decimals);
+        self
+    }
+
+    /// Set an exact number of decimals to display.
+    /// Values will also be rounded to this number of decimals.
+    /// Normally you don't need to pick a precision, as the slider will intelligently pick a precision for you.
+    /// Regardless of precision the slider will use "smart aim" to help the user select nice, round values.
+    pub fn fixed_decimals(mut self, num_decimals: usize) -> Self {
+        self.min_decimals = num_decimals;
+        self.max_decimals = Some(num_decimals);
+        self
+    }
 }
 
 impl<'a> Widget for DragValue<'a> {
@@ -113,11 +140,16 @@ impl<'a> Widget for DragValue<'a> {
             range,
             prefix,
             suffix,
+            min_decimals,
+            max_decimals,
         } = self;
+
         let value = get(&mut value_function);
         let aim_rad = ui.input().physical_pixel_size(); // ui.input().aim_radius(); // TODO
-        let precision = (aim_rad / speed.abs()).log10().ceil().at_least(0.0) as usize;
-        let value_text = format_with_minimum_precision(value as f32, precision); //  TODO: full precision
+        let auto_decimals = (aim_rad / speed.abs()).log10().ceil().at_least(0.0) as usize;
+        let max_decimals = max_decimals.unwrap_or(auto_decimals + 2);
+        let auto_decimals = clamp(auto_decimals, min_decimals..=max_decimals);
+        let value_text = format_with_decimals_in_range(value, auto_decimals..=max_decimals);
 
         let kb_edit_id = ui.make_position_id().with("edit");
         let is_kb_editing = ui.memory().has_kb_focus(kb_edit_id);
@@ -146,7 +178,12 @@ impl<'a> Widget for DragValue<'a> {
                 .sense(Sense::click_and_drag())
                 .text_style(TextStyle::Monospace);
             let response = ui.add(button);
-            let response = response.on_hover_text("Drag to edit or click to enter a value");
+            let response = response.on_hover_text(format!(
+                " {}{}{}\nDrag to edit or click to enter a value.",
+                prefix,
+                value as f32, // Show full precision value on-hover. TODO: figure out f64 vs f32
+                suffix
+            ));
             if response.clicked {
                 ui.memory().request_kb_focus(kb_edit_id);
                 ui.memory().temp_edit_string = None; // Filled in next frame
@@ -156,7 +193,7 @@ impl<'a> Widget for DragValue<'a> {
                 let delta_value = speed * delta_points;
                 if delta_value != 0.0 {
                     let new_value = value + delta_value as f64;
-                    let new_value = round_to_precision(new_value, precision);
+                    let new_value = round_to_decimals(new_value, auto_decimals);
                     let new_value = clamp(new_value, range);
                     set(&mut value_function, new_value);
                     // TODO: To make use or `smart_aim` for `DragValue` we need to store some state somewhere,
diff --git a/egui/src/widgets/slider.rs b/egui/src/widgets/slider.rs
index afe523fd..86b6c660 100644
--- a/egui/src/widgets/slider.rs
+++ b/egui/src/widgets/slider.rs
@@ -44,8 +44,9 @@ pub struct Slider<'a> {
     smart_aim: bool,
     // TODO: label: Option<Label>
     text: Option<String>,
-    precision: Option<usize>,
     text_color: Option<Srgba>,
+    min_decimals: usize,
+    max_decimals: Option<usize>,
 }
 
 impl<'a> Slider<'a> {
@@ -62,79 +63,68 @@ impl<'a> Slider<'a> {
             },
             smart_aim: true,
             text: None,
-            precision: None,
             text_color: None,
+            min_decimals: 0,
+            max_decimals: None,
         }
     }
 
     pub fn f32(value: &'a mut f32, range: RangeInclusive<f32>) -> Self {
-        Self {
-            ..Self::from_get_set(to_f64_range(range), move |v: Option<f64>| {
-                if let Some(v) = v {
-                    *value = v as f32
-                }
-                *value as f64
-            })
-        }
+        Self::from_get_set(to_f64_range(range), move |v: Option<f64>| {
+            if let Some(v) = v {
+                *value = v as f32
+            }
+            *value as f64
+        })
     }
 
     pub fn f64(value: &'a mut f64, range: RangeInclusive<f64>) -> Self {
-        Self {
-            ..Self::from_get_set(to_f64_range(range), move |v: Option<f64>| {
-                if let Some(v) = v {
-                    *value = v
-                }
-                *value
-            })
-        }
+        Self::from_get_set(to_f64_range(range), move |v: Option<f64>| {
+            if let Some(v) = v {
+                *value = v
+            }
+            *value
+        })
     }
 
     pub fn u8(value: &'a mut u8, range: RangeInclusive<u8>) -> Self {
-        Self {
-            ..Self::from_get_set(to_f64_range(range), move |v: Option<f64>| {
-                if let Some(v) = v {
-                    *value = v.round() as u8
-                }
-                *value as f64
-            })
-        }
+        Self::from_get_set(to_f64_range(range), move |v: Option<f64>| {
+            if let Some(v) = v {
+                *value = v.round() as u8
+            }
+            *value as f64
+        })
         .integer()
     }
 
     pub fn i32(value: &'a mut i32, range: RangeInclusive<i32>) -> Self {
-        Self {
-            ..Self::from_get_set(to_f64_range(range), move |v: Option<f64>| {
-                if let Some(v) = v {
-                    *value = v.round() as i32
-                }
-                *value as f64
-            })
-        }
+        Self::from_get_set(to_f64_range(range), move |v: Option<f64>| {
+            if let Some(v) = v {
+                *value = v.round() as i32
+            }
+            *value as f64
+        })
         .integer()
     }
 
     pub fn u32(value: &'a mut u32, range: RangeInclusive<u32>) -> Self {
-        Self {
-            ..Self::from_get_set(to_f64_range(range), move |v: Option<f64>| {
-                if let Some(v) = v {
-                    *value = v.round() as u32
-                }
-                *value as f64
-            })
-        }
+        Self::from_get_set(to_f64_range(range), move |v: Option<f64>| {
+            if let Some(v) = v {
+                *value = v.round() as u32
+            }
+            *value as f64
+        })
         .integer()
     }
 
     pub fn usize(value: &'a mut usize, range: RangeInclusive<usize>) -> Self {
         let range = (*range.start() as f64)..=(*range.end() as f64);
-        Self {
-            ..Self::from_get_set(range, move |v: Option<f64>| {
-                if let Some(v) = v {
-                    *value = v.round() as usize
-                }
-                *value as f64
-            })
-        }
+        Self::from_get_set(range, move |v: Option<f64>| {
+            if let Some(v) = v {
+                *value = v.round() as usize
+            }
+            *value as f64
+        })
         .integer()
     }
 
@@ -172,12 +162,37 @@ impl<'a> Slider<'a> {
         self
     }
 
-    /// Precision (number of decimals) used when displaying the value.
-    /// Values will also be rounded to this precision.
+    #[deprecated = "Use fixed_decimals instea"]
+    pub fn precision(self, precision: usize) -> Self {
+        self.max_decimals(precision)
+    }
+
+    // TODO: we should also have a "min precision".
+    /// Set a minimum number of decimals to display.
     /// Normally you don't need to pick a precision, as the slider will intelligently pick a precision for you.
     /// Regardless of precision the slider will use "smart aim" to help the user select nice, round values.
-    pub fn precision(mut self, precision: usize) -> Self {
-        self.precision = Some(precision);
+    pub fn min_decimals(mut self, min_decimals: usize) -> Self {
+        self.min_decimals = min_decimals;
+        self
+    }
+
+    // TODO: we should also have a "max precision".
+    /// Set a maximum number of decimals to display.
+    /// Values will also be rounded to this number of decimals.
+    /// Normally you don't need to pick a precision, as the slider will intelligently pick a precision for you.
+    /// Regardless of precision the slider will use "smart aim" to help the user select nice, round values.
+    pub fn max_decimals(mut self, max_decimals: usize) -> Self {
+        self.max_decimals = Some(max_decimals);
+        self
+    }
+
+    /// Set an exact number of decimals to display.
+    /// Values will also be rounded to this number of decimals.
+    /// Normally you don't need to pick a precision, as the slider will intelligently pick a precision for you.
+    /// Regardless of precision the slider will use "smart aim" to help the user select nice, round values.
+    pub fn fixed_decimals(mut self, num_decimals: usize) -> Self {
+        self.min_decimals = num_decimals;
+        self.max_decimals = Some(num_decimals);
         self
     }
 
@@ -185,7 +200,7 @@ impl<'a> Slider<'a> {
     /// If you use one of the integer constructors (e.g. `Slider::i32`) this is called for you,
     /// but if you want to have a slider for picking integer values in an `Slider::f64`, use this.
     pub fn integer(self) -> Self {
-        self.precision(0).smallest_positive(1.0)
+        self.fixed_decimals(0).smallest_positive(1.0)
     }
 
     fn get_value(&mut self) -> f64 {
@@ -193,8 +208,8 @@ impl<'a> Slider<'a> {
     }
 
     fn set_value(&mut self, mut value: f64) {
-        if let Some(precision) = self.precision {
-            value = round_to_precision(value, precision);
+        if let Some(max_decimals) = self.max_decimals {
+            value = round_to_decimals(value, max_decimals);
         }
         set(&mut self.get_set_value, value);
     }
@@ -325,7 +340,10 @@ impl<'a> Slider<'a> {
                     .text_style(TextStyle::Monospace)
                     .text_color_opt(self.text_color),
             );
-            let response = response.on_hover_text("Click to enter a value");
+            let response = response.on_hover_text(format!(
+                "{}\nClick to enter a value.",
+                self.get_value() as f32 // Show full precision value on-hover. TODO: figure out f64 vs f32
+            ));
             // let response = ui.interact(response.rect, kb_edit_id, Sense::click());
             if response.clicked {
                 ui.memory().request_kb_focus(kb_edit_id);
@@ -337,23 +355,26 @@ impl<'a> Slider<'a> {
     fn format_value(&mut self, aim_radius: f32, x_range: RangeInclusive<f32>) -> String {
         let value = self.get_value();
 
-        if let Some(precision) = self.precision {
-            format_with_minimum_precision(value as f32, precision)
+        // pick precision based upon how much moving the slider would change the value:
+        let value_from_x = |x: f32| self.value_from_x(x, x_range.clone());
+        let x_from_value = |value: f64| self.x_from_value(value, x_range.clone());
+        let left_value = value_from_x(x_from_value(value) - aim_radius);
+        let right_value = value_from_x(x_from_value(value) + aim_radius);
+        let range = (left_value - right_value).abs();
+        let auto_decimals = ((-range.log10()).ceil().at_least(0.0) as usize).at_most(16);
+        let min_decimals = self.min_decimals;
+        let max_decimals = self.max_decimals.unwrap_or(auto_decimals + 2);
+
+        let auto_decimals = clamp(auto_decimals, min_decimals..=max_decimals);
+
+        if min_decimals == max_decimals {
+            format_with_minimum_decimals(value, max_decimals)
         } else if value == 0.0 {
             "0".to_owned()
+        } else if range == 0.0 {
+            value.to_string()
         } else {
-            // pick precision based upon how much moving the slider would change the value:
-            let value_from_x = |x: f32| self.value_from_x(x, x_range.clone());
-            let x_from_value = |value: f64| self.x_from_value(value, x_range.clone());
-            let left_value = value_from_x(x_from_value(value) - aim_radius);
-            let right_value = value_from_x(x_from_value(value) + aim_radius);
-            let range = (left_value - right_value).abs();
-            if range == 0.0 {
-                value.to_string()
-            } else {
-                let precision = ((-range.log10()).ceil().at_least(0.0) as usize).at_most(16);
-                format_with_minimum_precision(value as f32, precision)
-            }
+            format_with_decimals_in_range(value, auto_decimals..=max_decimals)
         }
     }
 }