What pathping Does
pathping combines traceroute and ping into a single command. It first maps the route to the destination (like traceroute), then spends several minutes sending repeated probes to each hop along that route to estimate packet loss. This gives you a loss percentage per hop rather than a single end-to-end packet loss number from ping.
Basic Syntax and Flags
pathping 1.1.1.1
pathping speedtesthq.com
pathping -n 8.8.8.8 # skip DNS lookups for faster output
pathping -h 30 8.8.8.8 # trace up to 30 hops (default is 30)
pathping -p 250 8.8.8.8 # 250ms between pings per hop
pathping -q 200 8.8.8.8 # send 200 queries per hop (default 100)Flag Reference
| Flag | Meaning | When to Use |
|---|---|---|
-n | No reverse DNS lookup — show IP addresses only | Faster output; skip hostname resolution delays |
-h <hops> | Maximum number of hops to trace | Reduce to 15–20 if the destination is close |
-p <ms> | Milliseconds between probes per hop | Increase to reduce network load during the test |
-q <queries> | Number of probes per hop | Increase for higher-confidence loss estimates |
-w <ms> | Timeout in milliseconds per probe | Increase on high-latency paths (satellite, international) |
Understanding pathping Output
pathping output has two phases:
- Route discovery phase (first 10–15 seconds): lists each hop like traceroute, showing IP addresses and hostnames
- Statistics phase (2–5 minutes): probes each hop repeatedly and then prints a loss table
The statistics table has two loss columns for each hop:
- This Node/Link: loss attributed to the link between the previous hop and this hop
- Lost/Sent: cumulative loss from your computer to this hop
How to Read Loss Results Correctly
| Pattern | Meaning | Action |
|---|---|---|
| One middle hop shows 100% loss; all later hops are clean | That router deprioritizes ICMP to itself — normal behavior | Ignore — not a real packet loss problem |
| Loss starts at one hop and stays high through every subsequent hop | Real loss — problem is at or before that hop | Investigate that hop and report to ISP if it is in their network |
| Loss at the final destination only | Destination server is deprioritizing ICMP or rate-limiting | Verify with a TCP tool (nc, curl) — may not indicate real loss |
| Consistent loss across all hops including gateway | Local problem — Wi-Fi, LAN, router, or modem | Test on Ethernet; isolate to local network |
pathping vs traceroute vs MTR
| Tool | OS | Speed | Loss Measurement | Best For |
|---|---|---|---|---|
| pathping | Windows only | Slow (2–5 min) | Per-hop, over time | ISP tickets; no-install Windows diagnosis |
| traceroute / tracert | All | Fast (5–15 sec) | No — only latency per hop | Quick route mapping; finding where latency increases |
| MTR | Linux/macOS (Windows via WinMTR) | Continuous | Per-hop, live updating | Best combined view of path and loss in real time |
| ping (long run) | All | Any duration | End-to-end only | Confirming loss exists; monitoring over time |
When to Use pathping
- Intermittent lag or loss on Windows where you want built-in tooling
- Preparing evidence for an ISP support ticket — pathping output is easy to copy and paste
- When you want per-hop loss data but cannot install MTR or WinMTR
- Verifying whether loss is in your local network, your ISP, or beyond
Frequently Asked Questions
Is pathping better than traceroute?
They answer different questions. traceroute maps the path quickly and shows latency per hop — it takes seconds. pathping spends minutes probing each hop to estimate packet loss percentage — useful when you already know where latency increases but want to confirm whether loss is occurring at a specific point along the path.
Why does pathping take so long?
It sends 100 probes to each hop by default (adjustable with -q) and waits between probes. With 15+ hops, this adds up to several minutes. The slow speed is what makes the loss estimate statistically meaningful — a single dropped probe would be noise; 100 probes gives you a reliable percentage.
Does pathping work on macOS or Linux?
No. pathping is Windows-only. On macOS and Linux, use mtr (or WinMTR on Windows) for live per-hop loss statistics. MTR provides a continuously updating display that is more readable than pathping's output format and works across platforms.
Why does an intermediate hop show 100% loss when later hops are fine?
Many routers deprioritize or rate-limit ICMP packets addressed to themselves to protect against ping floods. They forward transit traffic normally but do not spend CPU time replying to probes targeting their own interface. If a hop shows 100% loss but every hop after it is clean, the traffic is flowing through — that router is simply not responding to the ICMP probes. This is expected behavior, not a real network problem.