diff --git a/src/arithmetic/mod.rs b/src/arithmetic/mod.rs
index 07bcf18..e9a50cf 100644
--- a/src/arithmetic/mod.rs
+++ b/src/arithmetic/mod.rs
@@ -7,6 +7,7 @@ pub(crate) mod addition;
 pub(crate) mod sqrt;
 pub(crate) mod cbrt;
 pub(crate) mod inverse;
+pub(crate) mod pow;
 
 /// Return 10^pow
 ///
diff --git a/src/arithmetic/pow.rs b/src/arithmetic/pow.rs
new file mode 100644
index 0000000..969f481
--- /dev/null
+++ b/src/arithmetic/pow.rs
@@ -0,0 +1,157 @@
+//! pow implementation
+
+use stdlib::num::NonZeroU64;
+
+use crate::*;
+
+/// Compute bd**exp using exponentiation by squaring algorithm, while maintaining the
+/// precision specified in ctx (the number of digits would otherwise explode).
+///
+/// Algorithm comes from https://en.wikipedia.org/wiki/Exponentiation_by_squaring
+pub(crate) fn impl_pow_with_context(bd: &BigDecimal, exp: i64, ctx: &Context) -> BigDecimal {
+    if exp == 0 {
+        return 1.into();
+    }
+
+    // When performing a multiplication between 2 numbers, we may lose up to 2 digits
+    // of precision.
+    // "Proof": https://github.com/akubera/bigdecimal-rs/issues/147#issuecomment-2793431202
+    const MARGIN_PER_MUL: u64 = 2;
+    // When doing many multiplication, we still introduce additional errors, add 1 more digit
+    // per 10 multiplications.
+    const MUL_PER_MARGIN_EXTRA: u64 = 10;
+
+    fn trim_precision(bd: BigDecimal, ctx: &Context, margin: u64) -> BigDecimal {
+        let prec = ctx.precision().get() + margin;
+        if bd.digits() > prec {
+            bd.with_precision_round(NonZeroU64::new(prec).unwrap(), ctx.rounding_mode())
+        } else {
+            bd
+        }
+    }
+
+    // Count the number of multiplications we're going to perform, one per "1" binary digit
+    // in exp, and the number of times we can divide exp by 2.
+    let mut n = exp.unsigned_abs();
+    // Note: 63 - n.leading_zeros() == n.ilog2, but that's only available in recent Rust versions.
+    let muls = (n.count_ones() + (63 - n.leading_zeros()) - 1) as u64;
+    // Note: div_ceil would be nice to use here, but only available in recent Rust versions.
+    let margin_extra = (muls + MUL_PER_MARGIN_EXTRA / 2) / MUL_PER_MARGIN_EXTRA;
+    let mut margin = margin_extra + MARGIN_PER_MUL * muls;
+
+    let mut bd_y: BigDecimal = 1.into();
+    let mut bd_x = if exp >= 0 {
+        bd.clone()
+    } else {
+        bd.inverse_with_context(&ctx.with_precision(
+            NonZeroU64::new(ctx.precision().get() + margin + MARGIN_PER_MUL).unwrap(),
+        ))
+    };
+
+    while n > 1 {
+        if n % 2 == 1 {
+            bd_y = trim_precision(&bd_x * bd_y, ctx, margin);
+            margin -= MARGIN_PER_MUL;
+            n -= 1;
+        }
+        bd_x = trim_precision(bd_x.square(), ctx, margin);
+        margin -= MARGIN_PER_MUL;
+        n /= 2;
+    }
+    debug_assert_eq!(margin, margin_extra);
+
+    trim_precision(bd_x * bd_y, ctx, 0)
+}
+
+#[cfg(test)]
+mod test {
+    use super::*;
+    use stdlib::num::NonZeroU64;
+
+    #[cfg(not(feature = "std"))]
+    macro_rules! println {
+        ( $( $x:expr ),* ) => {}
+    }
+
+    // Test that the 2 numbers are the same, assuming precision in ctx.
+    fn compare(bd: BigDecimal, bd_expected: BigDecimal, ctx: &Context) {
+        let bd_expected_round = bd_expected.with_precision_round(ctx.precision(), ctx.rounding_mode());
+        println!("100d  0123456789012345678901234567890123456789012345678901234567890123456789012345678901234567890123456789");
+        println!("exp   {}", bd_expected);
+        println!("val   {}", bd);
+        println!("exprd {}", bd_expected_round);
+
+        assert_eq!(bd, bd_expected_round);
+    }
+
+    macro_rules! impl_case {
+        ($name:ident: $start:expr, $exp:literal => $expected:literal) => {
+            #[test]
+            fn $name() {
+                let start = BigDecimal::from($start);
+                let exp = $exp;
+                let expected = $expected;
+                let ctx = Context::default();
+
+                println!("Compute {}**{}", start, exp);
+
+                let bd = start.pow_with_context(exp, &ctx);
+                let bd_expected = BigDecimal::from_str(expected).unwrap();
+
+                compare(bd, bd_expected, &ctx);
+            }
+        };
+    }
+
+    mod pow_known {
+        use super::*;
+
+        // Wolfram Alpha can get us to these precise values with a bit of log trickery, e.g.:
+        // 2**3000000000 = 10**log_10(2**3000000000) = 10**(3000000000 * log_10(2))
+        impl_case!(case_2_3000: 2, 3000 => "1.230231922161117176931558813276752514640713895736833715766118029160058800614672948775360067838593459582429649254051804908512884180898236823e903");
+        impl_case!(case_2_2048: 2, 2048 => "3.231700607131100730071487668866995196044410266971548403213034542752465513886789089319720141152291346368871796092189801949411955915049092109e616");
+        impl_case!(case_2_2001: 2, 2001 => "2.296261390548509048465666402355363968044635404177390400955285473651532522784740627713318972633012539836891929277974925546894237921726110662e602");
+        impl_case!(case_2_3000000000: 2, 3000000000 => "9.8162042336235053508313854078782835648991393286913072670026492205522618203568834202759669215027003865712110468405800021098042607617495e903089986");
+        // This works as 2 can be exactly inverted with only 1 digit (0.5).
+        impl_case!(case_0_5_30000000: BigDecimal::from(2).inverse(), 30000000 => "1.34921314623699835510360889355448887159595110457423959780496317685705095413905406464421931122265203166201415504288117880522818881981650e-9030900");
+        impl_case!(case_0_5_minus3000000000: BigDecimal::from(2).inverse(), -3000000000 => "9.8162042336235053508313854078782835648991393286913072670026492205522618203568834202759669215027003865712110468405800021098042607617495e903089986");
+        impl_case!(case_2_minus30000000: 2, -30000000 => "1.34921314623699835510360889355448887159595110457423959780496317685705095413905406464421931122265203166201415504288117880522818881981650e-9030900");
+        // This tests that the inverse operation carries enough digits to keep the precision.
+        impl_case!(case_3_minus30000000: 3, -30000000 => "2.2824965348198962029744520058679746159742143842721452620663907608967745444344346503448190515521985159162206416095535917875712100566195e-14313638");
+    }
+
+    macro_rules! impl_case {
+        ($name:ident: $start:expr, $exp:expr) => {
+            #[test]
+            fn $name() {
+                let start = BigDecimal::from_str($start).unwrap();
+                let exp = $exp.into();
+                let ctx = Context::new(NonZeroU64::new(50).unwrap(), RoundingMode::HalfEven);
+                let ctx_large = Context::new(NonZeroU64::new(500).unwrap(), RoundingMode::HalfEven);
+
+                println!("Compute {}**{}", start, exp);
+
+                let bd = start.pow_with_context(exp, &ctx);
+                let bd_expected = start.pow_with_context(exp, &ctx_large);
+
+                compare(bd, bd_expected, &ctx);
+            }
+        };
+    }
+
+    mod pow_misc {
+        use super::*;
+
+        // Test a few more misc values, checking that contexts with 50 or 500 precision
+        // get the same number, after scaling down the 500 precision result to 50.
+
+        impl_case!(case_misc_1: "-1.87421916986125215986", 3000912415i64);
+        impl_case!(case_misc_2: "230231922161117176931558813276752514640713895736833715766118029160058800614672948775360067838593459582429649254051804908512884180898236823e903", 1000151);
+        impl_case!(case_misc_3: "9.4215159218712961e-123", u32::MAX);
+        impl_case!(case_misc_4: "213", 3);
+        impl_case!(case_misc_5: "230231922161117176931558813276752514640713895736833715766118029160058800614672948775360067838593459582429649254051804908512884180898236823e903", -1000151);
+        impl_case!(case_misc_6: "9.4215159218712961e-123", i32::MIN);
+        // This test case fails without some extra margin (the number ends with 8.489 and gets rounded to 9 instead of 8)
+        impl_case!(case_misc_7: "-946773878.6364634037017822265625", 4294967295i64);
+    }
+}
diff --git a/src/lib.rs b/src/lib.rs
index 7f089e1..bb5324e 100644
--- a/src/lib.rs
+++ b/src/lib.rs
@@ -735,6 +735,32 @@ impl BigDecimal {
         }
     }
 
+    /// Raises the number to an integer power
+    ///
+    /// Uses default-precision, set from build time environment variable
+    //// `RUST_BIGDECIMAL_DEFAULT_PRECISION` (defaults to 100)
+    ///
+    /// ```
+    /// # use bigdecimal::BigDecimal;
+    /// let n: BigDecimal = 2.into();
+    /// assert_eq!(n.pow(3000000000), "9.816204233623505350831385407878283564899139328691307267002649220552261820356883420275966921502700387e903089986".parse().unwrap());
+    /// ```
+    #[inline]
+    pub fn pow(&self, exp: i64) -> BigDecimal {
+        self.pow_with_context(exp, &Context::default())
+    }
+
+    /// Raises the number to an integer power, using context for precision and rounding
+    ///
+    #[inline]
+    pub fn pow_with_context(&self, exp: i64, ctx: &Context) -> BigDecimal {
+        if self.is_zero() || self.is_one() {
+            return self.clone();
+        }
+
+        arithmetic::pow::impl_pow_with_context(&self, exp, ctx)
+    }
+
     /// Take the square root of the number
     ///
     /// Uses default-precision, set from build time environment variable