Plain Modbus/TCP is not secure by default. It carries no encryption, no authentication and no integrity checks, so it should never be exposed directly to untrusted networks. Where devices support it, we recommend adopting Modbus Security (mbaps), which adds TLS encapsulation and mutual X.509 authentication over the designated port 802. Where that is not yet possible, isolate the protocol behind firewalls, VPNs or brokered access immediately.
TL;DR:
- Test every production function code against each role, and recheck permissions after session resumption because cached decisions may persist longer than expected.
- For devices that cannot support TLS, inventory ports, deny unneeded traffic, and route remote access through authenticated VPNs, jump hosts, or managed gateways.
- Plan certificate issuance, renewal, expiry alerts, and revocation before rollout; pilot on noncritical devices and verify older controllers can handle modern cipher suites.
Table of Contents
- What Modbus Security (mbaps) is and how it differs from plain Modbus/TCP
- Why plain Modbus/TCP is vulnerable in modern OT environments
- How Modbus Security works in practice: mutual auth, RBAC and per-request authorisation
- Practical mitigations when you cannot immediately adopt Modbus Security
- Deploying Modbus Security at scale: PKI, certificate lifecycle and crypto agility
- Our perspective: a short audit checklist and supplier questions
- How we can help: audits, PKI planning and managed network services
- FAQ
- Sources
What Modbus Security (mbaps) is and how it differs from plain Modbus/TCP
Modbus Security wraps the existing Modbus Application Protocol data unit inside a TLS tunnel, much as HTTPS wraps HTTP. The underlying message format stays the same; what changes is the channel it travels through. According to the Modbus Organisation, the protocol mandates X.509v3 certificates for mutual authentication, so both client and server must prove their identity before any data moves, and it uses the IANA-assigned port 802 for secure traffic, instead of the legacy port 502.
The Modbus/TCP Security specification also extracts a role identifier from an X.509v3 certificate extension, which is used to apply access control per request rather than trusting every connected client equally.
Operationally, this means:
- Inventory every device’s current port usage before planning a migration to port 802.
- Confirm each endpoint supports TLS 1.2 or later with modern cipher suites.
- Treat the certificate as the access credential, not just a transport wrapper.
- Plan for coexistence, since legacy port 502 traffic will not disappear overnight.
Why plain Modbus/TCP is vulnerable in modern OT environments
Modbus was designed in an era when the operating assumption was a closed, trusted serial or physical network. That assumption no longer holds once devices sit on routable IP networks, connect to corporate IT, or get reached through remote vendor support links. The protocol itself has no concept of a password, no session encryption and no way to detect a tampered packet.
This creates a predictable set of attack vectors:
- Eavesdropping on unencrypted traffic to read process values and command sequences.
- Man-in-the-middle manipulation that alters register values in transit.
- Replay of previously captured write commands to force unwanted device states.
- Unauthorised write commands issued by any device that can reach the target port.
Peer-reviewed analysis of enhanced Modbus/TCP security points to the same gap: without added authentication and authorisation functions, the protocol remains open to tampering and interception. Default open ports, forgotten vendor remote-access channels and flat networks without segmentation all widen this exposure, particularly where OT and IT infrastructure have been merged for convenience rather than designed together.
How Modbus Security works in practice: mutual auth, RBAC and per-request authorisation
Once a client initiates a connection on port 802, the TLS handshake begins with both sides presenting certificates rather than the server alone, which is the key departure from typical web TLS. The process generally runs as follows:
- The client opens a TLS session and both parties exchange X.509v3 certificates.
- Each side validates the other’s certificate against a trusted certificate authority.
- The server extracts the role object identifier embedded in the client certificate.
- That role is mapped to a set of permitted function codes and registers.
- Each subsequent request is checked against that role’s rights before execution.
The Modbus/TCP Security specification confirms that role-to-rights mapping is left to the implementer, so the same certificate role may carry different permissions from one vendor’s device to another. This makes testing essential rather than optional.
Session resumption can reduce the handshake overhead on constrained devices, but it also means a cached role decision may persist longer than expected, so authorisation logic should be re-verified periodically rather than assumed for the life of a session.
Pro Tip: Test authorisation rules against every function code a device actually uses in production, not just the ones in the vendor’s default configuration.
Practical mitigations when you cannot immediately adopt Modbus Security
Many OT estates include devices that will never support TLS. For these, compensating controls do the work that the protocol cannot.
- Segment OT networks with deny-by-default access control lists, so only known source and destination pairs can communicate.
- Maintain a documented, current inventory of every protocol and port in use, which the NCSC’s operational technology guidance identifies as essential for accurate firewall rules and reduced attack surface.
- Broker data out of the OT zone rather than allowing inbound access: a DMZ-based historian, a proxy, or unidirectional replication keeps external systems from touching live controllers directly.
- Route any remote access through authenticated VPNs, jump hosts or managed vendor gateways rather than direct port exposure.
- Deploy next-generation firewalls with stateful and application-layer inspection, paired with intrusion detection tuned for industrial protocols, and baseline normal traffic so anomalies stand out.
- Where the risk profile demands strictly one-way flow, data diodes and cross-domain solutions remove the possibility of inbound compromise entirely.
NCSC guidance recommends treating the OT boundary itself as the primary control when devices cannot be upgraded, investing in boundary appliances that can be managed without touching legacy firmware. Our OT network segmentation approach follows the same logic: isolate first, then layer in protocol-level security as devices allow it.
Deploying Modbus Security at scale: PKI, certificate lifecycle and crypto agility
Mutual TLS only works if certificates are issued, renewed and revoked reliably, and this is where most mbaps rollouts stall. Teams typically choose between an existing enterprise certificate authority, a managed PKI service, or a vendor-supplied PKI tied to specific equipment, each with different trade-offs in control and ongoing cost.
A practical rollout sequence looks like this:
- Decide on a PKI model and document issuance and renewal procedures before buying hardware.
- Pilot certificate deployment on a small number of non-critical devices first.
- Build expiry monitoring and alerting so certificates never lapse unnoticed.
- Define a revocation process for compromised or decommissioned devices.
- Review hardware crypto capability on older controllers, since some cannot handle modern TLS cipher suites at all.
The Modbus/TCP Security specification sets out minimum TLS cipher expectations, and crypto agility matters here: algorithms considered safe today will eventually need replacing, so a workable upgrade and testing path should exist before wide rollout, not after.
Pro Tip: Budget for certificate lifecycle management as an ongoing operational task, not a one-off project milestone.
Our perspective: a short audit checklist and supplier questions

From the audits we run across manufacturing and logistics sites, the gap is rarely awareness, it is sequencing. A workable order is: inventory every Modbus-speaking device, isolate what cannot be secured yet, plan the PKI model, pilot mbaps on a non-critical cell, then monitor continuously.
When evaluating a vendor, we ask whether their devices support certificate-based authentication, which roles their RBAC model exposes, which TLS cipher suites they support, whether port 802 is implemented, and what their test procedure looks like before go-live. Our network infrastructure checklist captures the same inventory and compliance steps in more detail.
— Jacob
How we can help: audits, PKI planning and managed network services
We help industrial and manufacturing teams turn this guidance into a working project: a network audit that maps every Modbus device and port in use, PKI design support for mbaps rollouts, and managed services that keep certificates, monitoring and segmentation running day to day.
If you need a clear picture of where your OT network stands before you commit to a TLS migration, get in touch about a network audit and we will scope it with you.
FAQ
What are the downsides of using Modbus?
Classic Modbus/TCP has no built-in authentication, encryption or integrity checking, which leaves it open to eavesdropping, replay and unauthorised write commands on any network that can reach the device. These limitations are addressed by Modbus Security (mbaps), which adds TLS encapsulation and certificate-based authentication, or by compensating network controls where devices cannot be upgraded.
Is the TCP protocol safe?
TCP itself is a transport protocol and does not provide encryption or authentication on its own, so safety depends entirely on what runs on top of it. Plain Modbus/TCP inherits none of TCP’s reliability features as security, while Modbus Security adds TLS on top of the same TCP connection to protect it.
What does Modbus TCP mean?
Modbus TCP is a variant of the Modbus protocol that carries its standard register-based read and write commands over a TCP/IP network instead of a serial line. It typically runs on port 502 without encryption, which is why the Modbus Organisation introduced Modbus Security as a TLS-protected alternative on port 802.
How do I start securing a legacy Modbus TCP deployment?
Begin with a full inventory of devices, ports and data flows, then isolate anything that cannot be upgraded behind firewalls or a DMZ, following NCSC guidance on brokered connectivity. From there, pilot Modbus Security on a small, non-critical set of devices before wider rollout.
Sources
- Modbus Security: New Protocol to Improve Control System Security
- Enhanced Modbus/TCP security protocol (PMC/NIH article)
Recommended
- OT network segmentation for cybersecurity and OT teams
- Four Controls for MQTT Security in Enterprise IIoT
- Secure network design: proven examples for robust protection
- IoT security explained for IT decision-makers







