How many Zwave devices can I have and remain stable?

I did a quick search and couldn’t find an answer. I currently have 68 Zwave devices on my V3 hub and it is very stable. I know there is a breaking point, I just am not sure what it is on SmartThings. I have 3 more Zwave devices that I’d like to add at some point but don’t want to cross the stability line. I hit that limit at one time when I was on Iris which was ~70 more or less. The main problem was Zwave locks. I have since moved to Zigbee locks. Is there anyone out there with 70 to 80 Zwave devices on their system and it remains stable?

The short answer is none. :disappointed_relieved: The smartthings platform just isn’t stable at this time. The platform as a whole is going through a major transition: new app, new hub, new cloud platform. And they are specifically updating their Z wave implementation to be able to support the S2 security framework, and that has introduced a couple of glitches recently.

Hub Firmware Release Notes - 25.26

Smartthings has told us multiple times that their typical customer has 15 or fewer devices and never uses any custom code. So if you stay in that range, you’ll probably at least fit into what they test for.

I know that support on multiple occasions has told people that they are likely to have problems if they are running more than 30 zwave devices. I have no idea where they are getting that number from, and it hasn’t been posted officially.

There is a known problem if you try to add more than around 15 Z wave locks to one hub, but that seems to be specific to that device class. And it’s considered a bug, they are supposed to be looking at it.

There have been some community members who had over 150 zwave devices And they ran into all the usual platform and stability issues: sometimes things would work great for a while, then things which had been working great would stop working, then there would be a known bug that would get fixed, and repeat. Sometimes the workarounds or the fixes for known bugs were complicated by the fact that they had so many devices.

@johnconstantelo , @bamarayne , @anon36505037 , @ogiewon have all had large networks and may have more to add.

But really there’s just no telling.

You bet @JDRoberts!

At one point I had well over 100, perhaps close to 150. Over 100 were certainly switches/dimmers. They were a mix of Aeon, GE (old and new), Mimolites, and Zooz. Some provided metering data, some were just plain old on/off/dim. I also had an August lock that didn’t perform well either.

I would go for a long time without issues, and then like @anon36505037 said, the gods would wake up on the wrong side of the bed. Then it would be hell for several days, even over a month. Throw in ST issues, firmware updates, and backend changes we never knew about; and it became a full time job maintaining my ST environment.

I didn’t do zwave repairs very often (hardly at all in fact) because it did not do crap, in fact I think it made it worse. Every time I did one it would knock my FortrezZ Mimolites and water meter out, and I had to power cycle them to get them alive again.

ST is spending time improving their implementation of Z-wave, and improving the framework; but I am just not comfortable with their implementation and stability.

What I ended up deciding to do is put my time and effort into moving away from zwave and going to Zigbee.

The improvement in performance, speed, and stability has been incredible quite frankly. I have yet to experience any Zigbee issues that weren’t related to something I did wrong. When I’m done with my conversion, I’ll be down to just 19 (minimotes, zcombos, zooz outlets, and aeon micro switches and an HEM v1) zwave devices total out of 270 devices. Out of those remaining zwave devices, I’m actively looking for their zigbee counterparts.

What I have noticed during my conversion is an improvement in my zwave mesh performance. I’m below 40 zwave devices right now, and now my zwave repairs run in 15 minutes or less WITHOUT any issues, AND my FortrezZ water meter no longer requires a power cycle after a zwave repair. My Mimolites are long gone, but I bet those would be fine too.

My experience may not be typical, or something that someone else may experience; but knowing what I know now and what I’ve gone through, I would have focused on Zigbee devices from the very beginning and use Zwave where absolutely needed (like my HEM and Aeon micro controllers).

1 Like

I just wanted to emphasize that these instability issues are due to the SmartThings platform, not to zwave itself.

Over the years many people have had large installations of zwave devices using vera or Homeseer or Indigo Domotics without seeing any similar problems. But those platforms don’t have a cloud component.

Just sayin’… :wink:


Yup, exactly; hence me pointing out it’s ST’s implementation of Zwave. It just never felt right, especially when it came to all the capabilities left behind in their DTH’s that manufacturer’s built into their devices. ST did just the bare minimum to make devices work, and then that’s where Community developers stepped in. That’s my observation at least. Others may see it differently.


I can agree with all of the above. I’m currently running around 75 z wave and the rest of the system is zigbee and Wi-Fi for about 200 total.

My system was super stable for around a year and then the zwave gods found me a couple of months ago.

I changed nothing other than waking up one day so it really is all on the st end of the universe.


What make of Zigbee switches and dimmers are you using.

I’m just starting to build out, currently under 10 z-wave devices and trying to learn from other folks’ pain!

1 Like

Almost all have been GE switches and dimmers. They range from the older ZW4001 models to the newest zwave plus dimmers and fan controllers. Even as I’m typing this to you I have a GE zwave plus fan controller reporting as “Guest Bedroom Fan [40]: Not responding” in the IDE, but that could be related to the fact that I just replaced 4 GE zwave devices with their zigbee counterpart, and haven’t done a zwave repair yet. That may help resolve those messages because my mesh could use a clean up. I will say that as I’ve reduced my zwave device count, I’ve definitely seen a very noticeable improvement.


Remember, stability is a state of mind.

The less you mind, the more it’s stable.


Like @johnconstantelo, I have been moving away from Z-wave lately. I chose to replace all of my GE/Jasco switches and dimmers with Lutron Caseta. My lighting automations have never been more reliable. I am using Zigbee sensors (motion and contact) to trigger most of these lighting automations. The performance is very good. My wife has actually asked me to automate more of the house now that everything is running so stably. Of course, I should mention that I am now running on Hubitat for these automations, as I prefer local processing.


I was asked what I would add to the SmartThings Z wave implementation to make it competitive with other Z wave hubs.

In addition to overall stability and publishing a change log, these would be my top priorities, but in no particular order:

  1. A mapping utility

  2. support for controller shift

  3. Adopt Z wave tweaker as an official utility

  4. Full support for central scene commands, in particular the ability to create and manage scenes without having to use the button controller paradigm in a custom DTH. This should allow these devices to run locally in many cases. This becomes more important as more and more light switches are adopting central scene commands.

  5. fix whatever it is that causes support to tell people not to run Z wave repairs. :scream:

  6. A backup and restore utility (I recognize that this is a platform issue, not a zwave issue, but most of the competition have one.)


Spot on @JDRoberts. Hopefully ST will listen to those recommendations. They’ve all been desperately needed for quite some time.


+1. Z-Wave is very reliable when done properly.


So, who does it reliably?

Right now I’m in the “Samsung typical user size” of about 10 to 14 devices. But that’s mainly because I’m keeping my spend rate down by buying a couple of switches at a time. Long term I expect to have 50 to 60 smart switches. And maybe a handful of roller shades. Probably not much else as I find the window and door sensors too ugly.

While building stuff out is a fun hobby, I’m not wanting troubleshooting and maintenance to become a full time job!

I ran a Fibaro HC2 for about 4-5 years and it was rock solid. Admittedly, most of the Z-Wave devices were also Fibaro, but it played nicely with 3rd party stuff too for the most part. The thing is, fundamentally it’s a local device - the cloud is really only for remote access for the app and integrations. All the Lua scripts that run the automations are executed locally, and it’s insanely fast. They’ve clearly optimised it very well.

Unfortunately it totally and spectacularly failed on me! Not sure what happened, but it’s now a brick :frowning: I suspect it’s a memory of motherboard failure, which isn’t really their fault.

On the flip side, it’s not perfect either. Very fast and reliable, but doesn’t do Zigbee, doesn’t integrate with much else, app is functional but getting a bit dated now and lacks customisation…

However, my HC2 dying made me revisit Home Assistant and it’s awesome. I might get a replacement HC2 to run my Z-Wave network, and integrate it and everything else (SmartThings, Nest, Sonos, Unifi etc) into Home Asssitant. Best of everything that way.

1 Like