What must a seller do with over-collected California sales tax?
Edvin Givargis Published 6 minute read
The short answer
Return it to the customer or remit it to the state; California law forecloses the third option, keeping it, no matter how innocently the excess was collected. Sales tax in California is imposed on the retailer, and what the retailer adds to the invoice is reimbursement collected from the customer under contract. When the reimbursement collected exceeds the tax actually due, because the rate was configured wrong, a nontaxable line was taxed, or a district rate was applied that does not apply, the statute treats the excess as money the seller holds for someone else: it must either go back to the customer who paid it or go to the state with the seller's returns. The rule turns every billing error into a two-sided cleanup. Over-collected amounts create the refund-or-remit obligation, invoice by invoice and customer by customer. Under-collected amounts run the other way and are unambiguously the seller's problem, because the tax is the retailer's own liability whether or not the customer reimbursed it, and the state collects the shortfall from the seller regardless of what the invoice said. The cleanup that follows a discovered misconfiguration is therefore not one decision but a sort: each affected invoice lands in an over-collected pile, an under-collected pile, or a third pile that has nothing to do with tax at all, the customer who simply paid less than the invoice total, which is a receivable to chase, not a tax error to fix.
How the error happens, and how to size it
The classic generators of excess reimbursement are mechanical. A billing or project management system carries a default rate from another jurisdiction, the out-of-state or foreign seller applying its home consumption tax rate to United States invoices being the purest case; a statewide rate is applied where the destination's district rate is lower or where the sale is not taxable at all; a nontaxable service line inherits tax from a taxable line in a bundled configuration; or an exemption, resale, interstate shipment, electronic delivery, is missed in setup. The sizing exercise is an invoice-level workpaper: for each invoice in the affected window, what tax was charged, what tax was actually due at the correct destination rate on the correctly characterized lines, and the difference, signed. The workpaper's arithmetic will produce a net number, and the net is useful for understanding the size of the problem, but the statute does not net. The obligation runs to each customer who over-paid: customer A's excess cannot be kept because customer B was under-charged, and a seller that remits only the net has quietly retained part of what it collected from the over-charged customers. The honest structure treats the two piles separately, the full excess refunded or remitted, the full shortfall borne or recovered, and the netting confined to the internal memo that explains what the episode cost.
The two legal exits, and choosing between them
The customer-refund exit is the conservative one, and for sellers with ongoing relationships it is usually the right one: a credit note or refund to each over-charged customer, documented, with corrected invoices, closes the loop exactly where it opened and leaves the seller's returns needing only the accurate figures going forward. Its costs are administrative, finding the customers, processing small credits, explaining the episode, and it becomes genuinely hard where customers are numerous, small, or gone. The remit-to-state exit exists for exactly those cases: the seller pays the excess over with its returns, and the obligation is discharged because the money reached the sovereign that the invoice told the customer it was for. Mechanically, remitting a known dollar amount of excess through a return built on measures and rates takes a bottom-up computation, working backwards from the tax to be remitted through the applicable rate to an assumed measure that produces it, and the computation should be documented in the file so the return's apparent sales figure is explainable later. The choice between exits can be made pile by pile, refunds for the identifiable ongoing customers, remittance for the unreachable remainder, and the timing matters: the excess should move in the current or next filing period once quantified, because excess reimbursement sitting in the seller's account is the fact pattern that converts an innocent configuration error into something an auditor characterizes less charitably. One interaction to keep in view: the same principle governs in reverse when the seller later seeks money back from the state, because a refund of over-remitted tax is conditioned on the seller first returning the corresponding reimbursement to its customers; the state does not hand a seller a refund of money the seller collected from someone else.
The other two piles: shortfalls and short-payments
The under-collected pile belongs to the seller. California imposes the tax on the retailer, so the invoices that carried too little tax leave the seller owing the difference with its returns, and the only question is whether to absorb it or pursue the customers for the unbilled reimbursement, which is a contract and relationship decision, not a tax one; a seller may re-invoice for tax it should have charged, and whether that is worth doing depends on the amounts and the accounts. What the shortfall cannot do is hide inside the netting: the state is owed the full tax on the under-charged sales regardless of how much excess is being remitted on the over-charged ones. The third pile is the imposter: invoices where the tax was computed correctly but the customer paid less than the total, which surface in the same reconciliation because the cash received does not match the invoice. Those are accounts receivable, handled by collection, credit memo, or write-off under ordinary rules, and mixing them into the tax cleanup misstates both the tax obligation and the books. The reconciliation that sorts all three piles correctly, tax charged versus tax due versus cash received, is the deliverable that makes the rest of the cleanup mechanical.
Practice notes
The response sequence when a rate misconfiguration surfaces: freeze the error first, fix the billing configuration so the pile stops growing, then build the invoice-level workpaper for the affected window, sorting over-collections, under-collections, and short-payments into their separate piles. Choose the exit for the excess pile customer by customer, refund where relationships and logistics allow, remit the remainder, and move the money in the current cycle with the computation documented. Bear or bill the shortfall with eyes open, and route the short-payments to collections where they belong. Prospectively, the controls are the ones the companion articles keep repeating: destination-rate lookup wired into billing rather than a hard-coded rate, line-level taxability that follows the product taxonomy, and a periodic sample reconciliation of charged tax against computed tax, because the seller who finds its own error chooses among the exits above, while the seller whose auditor finds it has the same obligations plus interest, penalties, and an examiner's framing of why the excess sat in the seller's account.
This article states the law as of September 16, 2026
Statutes, rates, thresholds, and agency practice change, and a different set of facts can change the answer. Before acting on anything discussed above, contact G&G State Tax Group to confirm what has changed since this was written and how the rules apply to a specific situation.
G&G State Tax Group, LLC is a state and local tax advisory firm. The firm provides state and local tax consulting and representation in state and local tax controversies. G&G does not prepare or file tax returns, perform attest services, or provide bookkeeping, and is not a CPA firm.