There is no universal "correct" number of dependent claims. Most patent practitioners aim for a tiered structure of three to five dependent claims per independent claim, balancing the cost of additional filing fees against the need for fallback positions if the broadest claim is invalidated. The optimal count depends on the complexity of your invention, the specific jurisdiction’s fee schedule for extra claims, litigation strategy, and your goals during prosecution.
Why the Number of Dependent Claims Matters
Dependent claims are not just extra words; they are safety nets. An independent claim is broad and risky. If an examiner finds prior art that anticipates your broad claim, you lose that protection. Dependent claims narrow the scope by adding specific features, creating a hierarchy of protection. If claim 1 fails, claim 2 might survive. If claim 2 fails, claim 3 might survive.
However, every additional claim costs money. In many jurisdictions, you pay a base fee for a set number of claims (often 20) and then a per-claim fee for anything beyond that. Adding ten unnecessary dependent claims might cost you significant fees without adding real value. Conversely, having too few dependent claims leaves you with no fallback if your main claim is rejected. You want enough layers to survive a rejection but not so many that you are paying for redundancy.
Balancing Protection and Filing Costs
The trade-off is between breadth and cost. Broad claims cover more ground but are easier to invalidate. Narrow claims are harder to invalidate but cover less market share. You should add a dependent claim if it serves a strategic purpose, such as establishing a fallback position, clarifying a commercial embodiment, or providing a distinct feature that a competitor might try to design around.
Consider the cost structure. If your jurisdiction charges for every claim over a threshold, you must be ruthless. Do not add a dependent claim just because you can. Ask: "If this specific feature is removed, does the invention still work?" If yes, and if that feature is a likely target for competitors to omit or a useful fallback for prosecution, then it is a good candidate for a dependent claim. If the feature is essential to the invention, it likely belongs in the independent claim or a higher-level dependent claim.
Structuring a Smart Claim Hierarchy
A smart hierarchy moves from broad to specific in logical steps. Start with the independent claim that defines the core invention. Then, add dependent claims that each introduce one additional limitation. Avoid stacking too many limitations into a single dependent claim. If you combine five features into one claim, you make it very narrow. If you spread them out, you have more chances for one of them to be allowed.
For a software invention, this often means layering by component, then by protocol, then by specific implementation detail. This structure allows you to argue for allowance at different levels of specificity during prosecution. If the examiner rejects the broad protocol, you can argue that the specific implementation detail is novel.
Worked Example: Secure Login System
Let us look at a concrete example for a software invention: a secure login system that uses multi-factor authentication (MFA) with a time-based token.
Independent Claim 1 (Broad): A computer-implemented method for authenticating a user, comprising: receiving a user identifier; generating a temporary token based on a shared secret and a current timestamp; receiving a user-entered token; and granting access if the user-entered token matches the generated token.
Now, we add dependent claims to create fallback positions.
Dependent Claim 2 (Adds Specificity to Token Generation): The method of claim 1, wherein the shared secret is stored on a hardware security module (HSM) isolated from the main application server.
Why this works: This claim narrows the scope by specifying the hardware environment. If the examiner finds prior art for general HSMs, you still have the broader claim. If they find prior art for HSMs in this specific context, you still have the broader claim. But if the HSM aspect is your key novelty, this claim protects it.
Dependent Claim 3 (Adds Specificity to the Token): The method of claim 1, wherein the temporary token is a six-digit numeric code generated using a Time-based One-Time Password (TOTP) algorithm.
Why this works: This narrows the scope from "any token" to a specific standard. If the broad claim is rejected, this specific implementation might be allowed if the prior art does not explicitly teach this combination with the other features.
Dependent Claim 4 (Adds Specificity to the Validation): The method of claim 1, further comprising: storing a set of previously used tokens for a predetermined time window; and rejecting the user-entered token if it matches any token in the set of previously used tokens.
Why this works: This adds a replay attack protection feature. It is a distinct functional addition. If the examiner rejects the token generation, this claim focuses on the validation logic.
Dependent Claim 5 (Combines Limitations for Maximum Specificity): The method of claim 2, wherein the HSM generates the token using a TOTP algorithm and the validation step includes checking against the set of previously used tokens.
Why this works: This is your narrowest position. It combines the HSM, TOTP, and replay protection. It is very hard to invalidate because it requires the examiner to find prior art that teaches all three specific elements together. It is your last line of defense.
Notice we have four dependent claims. This is a typical, balanced number. It provides four distinct fallback positions (Claims 2, 3, 4, and 5) without being excessive.
How AI Tools Optimize Claim Counts
Determining the right number of layers requires seeing the invention from multiple angles. Humans often miss obvious fallback positions or over-claim on trivial features. AI tools can analyze your technical description and suggest a structured hierarchy that balances breadth and specificity automatically.
ClaimGen uses a Smart Claim Hierarchy feature to auto-structure independent and dependent claims. It takes your raw technical notes and generates a draft set where the dependent claims logically branch from the independent claim. The goal is to get a high-quality initial draft that already has a sensible claim count, reducing the hours your attorney needs to spend restructuring the claims.
Common Mistakes in Dependent Claim Drafting
The most common mistake is the "kitchen sink" approach. This is when you dump every possible feature into one dependent claim. This creates a claim that is so narrow it is almost useless because competitors can easily design around it by removing one minor feature. Keep each dependent claim focused on one or two related limitations.
Another mistake is circular dependencies. Ensure each dependent claim clearly references a previous claim. Do not have Claim 5 depend on Claim 4, which depends on Claim 3, which depends on Claim 1, if Claim 3 is actually a better parent for Claim 5. A flat hierarchy is often easier to defend than a deep, nested one.
Finally, do not ignore the cost. If you are filing in a jurisdiction with high per-claim fees, consider whether a feature is worth the extra cost. If a feature is obvious or already well-known in the industry, it may not be worth a dependent claim. Save those claims for features that are truly novel or non-obvious.
When in doubt, draft the claims, then have a qualified patent attorney review the hierarchy. They can tell you if your number of claims is appropriate for your specific budget and legal strategy.