diff --git a/src/netcheck/stats.py b/src/netcheck/stats.py
new file mode 100644
index 0000000..ec64c28
--- /dev/null
+++ b/src/netcheck/stats.py
@@ -0,0 +1,53 @@
+from __future__ import annotations
+
+from typing import NamedTuple
+
+
+class RttStats(NamedTuple):
+    sent: int
+    received: int
+    loss_pct: float
+    min_ms: float | None
+    avg_ms: float | None
+    max_ms: float | None
+    mdev_ms: float | None
+    jitter_ms: float | None
+
+
+def rtt_stats(samples: list[float | None]) -> RttStats:
+    sent = len(samples)
+    got = [s for s in samples if s is not None]
+    received = len(got)
+    loss_pct = 0.0 if sent == 0 else round(100.0 * (sent - received) / sent, 3)
+    if not got:
+        return RttStats(sent, 0, loss_pct, None, None, None, None, None)
+    avg = sum(got) / received
+    mdev = sum(abs(v - avg) for v in got) / received
+    deltas = [
+        abs(b - a)
+        for a, b in zip(samples, samples[1:])
+        if a is not None and b is not None
+    ]
+    jitter = sum(deltas) / len(deltas) if deltas else 0.0
+    return RttStats(
+        sent=sent,
+        received=received,
+        loss_pct=loss_pct,
+        min_ms=round(min(got), 3),
+        avg_ms=round(avg, 3),
+        max_ms=round(max(got), 3),
+        mdev_ms=round(mdev, 3),
+        jitter_ms=round(jitter, 3),
+    )
+
+
+def percentile(values: list[float], p: float) -> float:
+    if not values:
+        return 0.0
+    ordered = sorted(values)
+    if len(ordered) == 1:
+        return ordered[0]
+    rank = (p / 100.0) * (len(ordered) - 1)
+    low = int(rank)
+    high = min(low + 1, len(ordered) - 1)
+    return ordered[low] + (ordered[high] - ordered[low]) * (rank - low)
diff --git a/tests/test_stats.py b/tests/test_stats.py
new file mode 100644
index 0000000..2557a16
--- /dev/null
+++ b/tests/test_stats.py
@@ -0,0 +1,80 @@
+from __future__ import annotations
+
+import math
+
+import pytest
+
+from netcheck.stats import percentile, rtt_stats
+
+
+def test_all_samples_present():
+    s = rtt_stats([10.0, 12.0, 14.0, 16.0])
+    assert s.sent == 4
+    assert s.received == 4
+    assert s.loss_pct == 0.0
+    assert s.min_ms == 10.0
+    assert s.max_ms == 16.0
+    assert s.avg_ms == 13.0
+    assert s.mdev_ms == pytest.approx(2.0)
+    assert s.jitter_ms == pytest.approx(2.0)
+
+
+def test_loss_is_the_share_of_missing_samples():
+    s = rtt_stats([10.0, None, None, 20.0])
+    assert s.sent == 4
+    assert s.received == 2
+    assert s.loss_pct == 50.0
+    assert s.avg_ms == 15.0
+
+
+def test_jitter_ignores_gaps_and_uses_consecutive_received_pairs():
+    # Consecutive received pairs are (10,20) and (20,26): mean |diff| = (10+6)/2 = 8.
+    s = rtt_stats([10.0, 20.0, None, 20.0, 26.0])
+    assert s.jitter_ms == pytest.approx(8.0)
+
+
+def test_all_timeouts_yield_full_loss_and_no_timing_values():
+    s = rtt_stats([None, None, None])
+    assert s.sent == 3
+    assert s.received == 0
+    assert s.loss_pct == 100.0
+    assert s.min_ms is None
+    assert s.avg_ms is None
+    assert s.max_ms is None
+    assert s.mdev_ms is None
+    assert s.jitter_ms is None
+
+
+def test_single_sample_has_zero_deviation_and_zero_jitter():
+    s = rtt_stats([13.5])
+    assert s.received == 1
+    assert s.min_ms == s.avg_ms == s.max_ms == 13.5
+    assert s.mdev_ms == 0.0
+    assert s.jitter_ms == 0.0
+
+
+def test_empty_input_is_not_a_division_by_zero():
+    s = rtt_stats([])
+    assert s.sent == 0
+    assert s.received == 0
+    assert s.loss_pct == 0.0
+    assert s.avg_ms is None
+
+
+def test_every_stat_is_finite_so_json_serialization_cannot_break():
+    for samples in ([], [None], [1.0], [1.0, None, 3.0]):
+        for value in rtt_stats(samples):
+            assert value is None or math.isfinite(value)
+
+
+def test_percentile_interpolates_between_ranks():
+    values = [1.0, 2.0, 3.0, 4.0]
+    assert percentile(values, 0.0) == 1.0
+    assert percentile(values, 100.0) == 4.0
+    assert percentile(values, 50.0) == pytest.approx(2.5)
+    assert percentile(values, 90.0) == pytest.approx(3.7)
+
+
+def test_percentile_handles_single_and_empty_inputs():
+    assert percentile([7.0], 90.0) == 7.0
+    assert percentile([], 90.0) == 0.0