Threat Modeling Methodologies Decoded: STRIDE, PASTA, LINDDUN & Attack Trees
Choosing the right lens for software engineering, business risk, privacy & deep exploit paths
“A carpenter doesn’t argue about whether a hammer is better than a saw. Pick the threat modeling methodology that matches the question you need answered.”
Walk into an AppSec conference and ask "Which threat modeling methodology is best?" and you will start a 45-minute philosophical debate. One person swears by STRIDE, another insists on 7-stage PASTA, and a privacy engineer chimes in with LINDDUN.
Here is the secret: they are not competitors—they are different lenses. Looking at a microservice through STRIDE reveals broken authentication and tampering. Looking at that same microservice through LINDDUN reveals that your analytics pipeline accidentally de-anonymizes minors in Europe.
Knowing how each methodology thinks lets you pick the sharpest tool in 30 seconds instead of forcing a heavyweight framework onto a two-day sprint.
Comparing the Big Four: STRIDE, PASTA, LINDDUN & Attack Trees
STRIDE (Spoofing, Tampering, Repudiation, Information Disclosure, Denial of Service, Elevation of Privilege) is software-centric. Each letter maps directly to a security property engineers understand: Authentication (S), Integrity (T), Non-repudiation (R), Confidentiality (I), Availability (D), and Authorization (E).
PASTA (Process for Attack Simulation and Threat Analysis) is risk-centric across 7 stages—starting with business objectives and regulatory impact before decomposing tech and simulating attacks. LINDDUN is privacy-centric (Linkability, Identifiability, Non-repudiation, Detectability, Disclosure, Unawareness, Non-compliance), essential for GDPR/CCPA systems.
Attack Trees are goal-centric: place an attacker goal at the root ("Steal Customer Signing Keys") and branch downward with OR nodes (alternative paths) and AND nodes (conditions that must all hold). Combining STRIDE for breadth with Attack Trees for Crown Jewels is a superpower.
Fast, repeatable, and maps directly to concrete technical controls on Data Flow Diagrams.
Connects threat scenarios directly to financial impact and business priorities across 7 structured stages.
Catches tracking, linkability, and metadata leakage that classic security reviews miss.
Decomposes high-value targets into concrete AND/OR exploit prerequisites.
Apply S/T/R/I/D/E systematically to every Process, Data Store, and Data Flow.
Crucial when "encrypted" data can still be correlated via timestamps or IP metadata.
Every leaf node becomes a concrete test case or security ticket.
Case Study: Why Classic Security Missed a Privacy Breach (Until LINDDUN)
A fitness app encrypted all GPS coordinates in transit (TLS 1.3) and at rest (AES-256) and enforced strong OAuth tokens. By STRIDE alone, the public leaderboard API looked clean—until researchers realized public pace-per-minute + start timestamps allowed anyone to infer home addresses.
Transport & Auth Passed
TLS, OAuth, and database encryption satisfied classic confidentiality and spoofing checks.
Linkability & Identifiability Fired
LINDDUN asks: "Can an external observer link anonymized records back to a real person?" Public start/end zones made home locations trivial to triangulate.
Privacy Zones & Spatial Fuzzing
Trimming the first and last 300 meters of public routes eliminated the linkability risk at zero performance cost.
There Are No Dumb Questions
Candid answers to the skeptical questions every security engineer asks first.
OR means the attacker only needs ONE of the child steps to succeed. AND means the attacker must achieve ALL child steps—so breaking just one AND leaf blocks that entire branch!STRIDE vs. PASTA in Modern Engineering Orgs
“Without tying threats to business objectives first, engineers waste time fixing low-impact theoretical bugs!”
“True—so we tag Crown Jewels for business impact, then run fast STRIDE on every Data Flow. Best of both worlds without a 3-week workshop!”
Sharpen Your Pencil · 60-Second Brain Challenge
You are reviewing a healthcare analytics portal that strips patient names from exported CSVs, but leaves exact birthdates, 5-digit ZIP codes, and visit timestamps intact. Which methodology specifically catches this re-identification risk?
- ▸STRIDE is the fastest, most developer-friendly lens for catching technical flaws on Data Flow Diagrams.
- ▸PASTA connects technical threats to business objectives and financial impact across 7 stages.
- ▸LINDDUN uncovers privacy blind spots (Linkability, Identifiability, Detectability) that encryption alone cannot solve.
- ▸Attack Trees map concrete
AND/ORpaths to Crown Jewels—break oneANDprerequisite to neutralize the chain.
Practice running deterministic STRIDE checks on your own topology in NodeMotion’s Live Canvas →