diff --git a/src/bin/day3.rs b/src/bin/day3.rs
index 810c108..465f0e6 100644
--- a/src/bin/day3.rs
+++ b/src/bin/day3.rs
@@ -6,13 +6,24 @@ use std::collections::HashSet;
 use std::fs::File;
 use std::io::{BufRead, BufReader};
 
+// Map a character (indicating an "item") to its "priority".
+fn char_to_prio(c: char) -> u32 {
+    // Fortunately we can cast all chars to u32, which makes this a matter of some simple
+    // arithmetic.
+    match c {
+        'a'..='z' => c as u32 - 'a' as u32 + 1,
+        'A'..='Z' => c as u32 - 'A' as u32 + 27,
+        _ => panic!("Got non-ascii alphabetic char!"),
+    }
+}
+
 fn main() -> StdError<()> {
     // Open the input file and split all the lines.
     let lines: Vec<_> = BufReader::new(File::open("input/day3.txt")?)
         .lines()
         .collect::<Result<_, _>>()?; // "Short circuit" any Errors in the iterator
 
-    let sum: u32 = lines
+    let part1: u32 = lines
         .iter()
         .map(|l| {
             assert!(l.len() % 2 == 0); // We won't be able to evenly split a odd-length line!
@@ -22,17 +33,35 @@ fn main() -> StdError<()> {
             HashSet::<char>::from_iter(one.chars())
                 // ... then calculate the intersection, giving us the "items" (chars) in common
                 .intersection(&HashSet::from_iter(two.chars()))
-                .map(|c| match c {
-                    // Now, we map our "items" to their "priorities".
-                    // Fortunately we can cast all chars to u32, which makes this a matter of some
-                    // simple arithmetic.
-                    'a'..='z' => (*c as u32) - 'a' as u32 + 1,
-                    'A'..='Z' => (*c as u32) - 'A' as u32 + 27,
-                    _ => panic!("Got non-ascii alphabetic char!"),
-                })
+                .map(|c| char_to_prio(*c))
                 .sum::<u32>() // Sum the priorities of items in common in each "rucksack"
         })
         .sum(); // Sum all the rucksacks' priorities
-    println!("Part 1: {}", sum);
+    println!("Part 1: {}", part1);
+
+    let part2: u32 = lines
+        // First, map our lines to hashsets of their chars
+        .iter()
+        .map(|l| HashSet::from_iter(l.chars()))
+        .collect::<Vec<_>>()
+        // Then split them into chunks of 3
+        .chunks(3)
+        .map(|c: &[HashSet<_>]| {
+            // And reduce each chunk of 3 into a single intersection of chars...
+            let common: Vec<char> = c
+                .into_iter()
+                .cloned()
+                .reduce(|acc, s| acc.intersection(&s).cloned().collect())
+                .unwrap()
+                .into_iter()
+                .collect();
+            // ... which should only be one!
+            assert!(common.len() == 1);
+            common[0]
+        })
+        .map(char_to_prio)
+        .sum();
+    println!("Part 2: {}", part2);
+
     Ok(())
 }