How to Dropship from China in 2026: Complete Guide

Dropshipping from China in 2026 still works, but not the way it did in 2020. The model is unchanged: you list products, a Chinese supplier ships each order to your customer, and you never touch inventory. What changed is the cost floor. The US ended duty-free treatment for low-value parcels from China in May 2025 and from all other countries in August 2025, so every shipment now carries import duty that has to sit inside your margin.

The survivors adjusted their math: duty baked into landed cost, products with margin to absorb it, and faster fulfillment. CN Ally helps sellers do exactly that, finding suppliers who deliver what they promise and checking quality before goods ship.

This guide covers how the model works now, the three platforms that matter, realistic shipping times, the new duty reality, product selection, common pitfalls, and when to stop dropshipping and buy in bulk.

How dropshipping from China actually works

The mechanics are simple. You set up an online store, import product listings from a Chinese supplier, and price them above your cost. When a customer buys, you forward the order to the supplier, pay the wholesale price plus shipping, and the supplier ships the item straight to the customer. You keep the difference. No warehouse, no packing, no inventory risk.

What the simple version leaves out is everything that makes the model succeed or fail: supplier processing time, international shipping speed, customs clearance, and what happens when the product arrives broken or never arrives at all. In dropshipping, you own the customer relationship while controlling none of the fulfillment, and the refund comes out of your account when either side fails.

Most beginners start on Shopify connected to AliExpress through an automation app. As volume grows, many graduate to a fulfillment platform like CJ Dropshipping, which warehouses products in the US or Europe, or to a private sourcing agent who buys from 1688 factories and handles QC and consolidation. Each step up costs more per unit but buys speed and reliability.

The duty change that rewrote the math

For years, Section 321 of the Tariff Act of 1930 let shipments valued at $800 or less enter the US without duty or a formal customs entry. Roughly 1.36 billion parcels used it in fiscal 2024, about 60% from China, per trade compliance reporting. Two executive orders dismantled it: EO 14256 ended it for China and Hong Kong effective May 2, 2025, and EO 14324 extended the suspension to all countries effective August 29, 2025. In June 2026, CBP published interim final rules making the suspension indefinite, and statute repeals the commercial exception effective July 1, 2027. (Full timeline at ghy.com) The EU is also introducing a flat EUR 3 per item duty on low-value imports up to EUR 150 starting July 1, 2026, per industry trade reporting.

The practical effect for your store is a new fixed cost per order that ranges from a few dollars to much more depending on the product's HS code. A $12 phone case that used to clear duty-free may now carry several dollars of duty. If your gross margin was $6, the business is gone. If it was $18, you're still alive. Every order now clears customs, with your supplier or forwarder handling the entry and you paying through shipping fees. Ship DDP (delivered duty paid) so the customer sees one price: surprise customs fees on delivery are the fastest route to chargebacks and one-star reviews.

Build the duty estimate into product research from day one. Ask your supplier or agent for the typical landed cost per unit on your route, and treat that number, not the product page price, as your cost of goods.

AliExpress: where almost everyone starts

AliExpress remains the default entry point, and for good reason. The catalog is enormous, prices are close to factory level, and there is no minimum order, so you can list a product tonight and test demand with real orders tomorrow. The official automation tool is DSers, free for up to 3,000 products on its starter plan, which handles bulk listing imports, automatic order placement, and stock and price syncing.

The tradeoff is variability. You are ordering from thousands of independent sellers, each with their own processing time, shipping method, and quality standards. A 90-day head-to-head test published in 2026 measured AliExpress standard shipping to US addresses at 17.4 days average, within a 7-to-25-day window, and a 6.8% refund rate. (Comparison at rightcosta.com) AliExpress Choice listings narrow delivery to 5 to 10 days for select products, but coverage is uneven.

Use AliExpress for what it's good at: cheap product testing. Order samples of everything before you list it, stick to sellers with long track records, and favor listings that ship via AliExpress Standard. Message the seller before scaling a winner to confirm stock and processing time. Understand the ceiling: inconsistent branding, no real QC, and long delivery times make AliExpress a testing ground, not a long-term fulfillment partner.

CJ Dropshipping: the scaling platform

CJ Dropshipping is not a marketplace. It's a fulfillment company with its own warehouses and a curated catalog, and that difference shows up in delivery times: a 90-day head-to-head test published in 2026 measured CJ at 11.5 days average to US addresses against 17.4 for AliExpress standard shipping. Products stocked in CJ's US, EU, UK, Canadian, or Australian warehouses can reach customers in days rather than weeks.

The platform is free to use; you pay for the product, shipping, and optional services. Unit prices usually run 10 to 20% above the equivalent AliExpress listing, since CJ folds warehousing, inspection, and its agent service into the price. What you get for the markup: custom sourcing (send them any product and they will find it), white-label fulfillment with custom packaging, in-house product photography, and a personal agent once your volume justifies one.

The sellers who do best on CJ are past the testing phase and scaling one or two winners. At that point, the cheapest price and biggest catalog matter less than the things that protect margin: faster delivery, fewer refunds, and branding that makes the product yours instead of interchangeable with ten other stores selling the same item.

1688: the factory-direct edge

1688 is Alibaba's domestic wholesale marketplace: built for the Chinese market, in Chinese, for Chinese buyers. It is also where many AliExpress products originate, typically 10 to 30% cheaper than their Alibaba equivalents according to wholesale sourcing guides, because sellers price for domestic buyers with no export markup.

That price gap is real, but it comes with friction that makes 1688 unsuitable as a direct dropshipping platform. The interface is Chinese-only. Most suppliers don't speak English, ship only domestically within China, and prefer Alipay or WeChat Pay through local rails. You can't list a 1688 product in your Shopify store and expect orders to fulfill themselves.

The workaround is a purchasing agent: they buy on your behalf, inspect goods, consolidate parcels from multiple suppliers, and ship internationally with tracking. The workable pattern is a hybrid: test on AliExpress or CJ, then move proven winners to a 1688 agent for better pricing, custom packaging, and a quality check before every shipment. This is the stage where a sourcing partner pays for itself, which is exactly the service CN Ally provides for sellers who have outgrown marketplace fulfillment.

Which platform fits your stage

The three options are a ladder, not competitors. Here's how they compare:

Factor · AliExpress (via DSers) · CJ Dropshipping · 1688 (via agent)

• Catalog size: 100M+ SKUs · ~1M curated · Millions, factory-direct
• US shipping time: 7–25 days (avg ~17) · 7–20 days (avg ~11.5) · Depends on agent, often 7–15 days
• US warehouse option: No · Yes · Sometimes, via agent stock
• Product cost: Lowest · ~10–20% higher · 10–30% below Alibaba equivalents
• Branding / private label: Limited · Strong (custom packaging, inserts) · Full, via agent
• Custom sourcing: No · Yes · Yes, effectively the whole site
• Automation: DSers (free tier) · Built-in app · Agent-dependent
• Language barrier: None · None · Significant without agent
• Best for: Testing products cheaply · Scaling winners with branding · Margin optimization at volume

Most successful stores climb this ladder in order. Testing on AliExpress keeps the cost of a failed product near zero. Moving winners to CJ buys speed and branding without managing inventory. Shifting proven, high-volume SKUs to 1688 through an agent recovers margin and adds a QC layer. Skipping straight to 1688 as a beginner usually wastes money on products that haven't proven they sell.

Picking products that survive the new math

Product selection matters more now, because the products that died are the ones whose margins were too thin to absorb duty. Run every candidate through this filter before you test it.

Sell price between $25 and $75. Below $25, duty and shipping eat the margin. Above $75, impulse conversion drops and returns get expensive. The sweet spot gives roughly 3x to 5x markup over landed cost.

Landed cost is the only cost that matters. Product price plus shipping plus duty plus transaction fees, per unit. A $9 product with $4 shipping and $3 duty is a $16 product. If you can't sell it for $40+, keep looking.

Light, small, and hard to break. Shipping is priced on weight and dimensions, and fragile goods generate refunds. Products under 500 grams that survive a drop test keep both freight and refund rates low.

Not trademarked, not regulated, not dangerous. Knockoff branded goods get your store shut down; supplements, medical claims, and uncertified electronics invite compliance problems. Check the platform's prohibited items list before you list.

Solves a visible problem or scratches a visible itch. Clear problem-solvers and clear novelty win. Generic commodities with no angle compete purely on price, and someone with a 1688 agent will always beat you there.

Check the competition's delivery promise. If five established stores sell the same item with 3-day delivery from a US warehouse and you offer 15-day shipping from China, you need a real differentiator: better creative, a bundle, or a lower price.

The pitfalls that kill dropshipping stores

Most failed stores don't fail because the model is broken. They fail on the same handful of preventable mistakes.

Promising delivery times you can't keep. A store that advertises "fast shipping" and delivers in 22 days earns disputes. State realistic windows, underpromise slightly, and use tracking updates to keep buyers informed. The refund you avoid is worth more than the conversion you lose to an honest delivery estimate.

Skipping samples. A sample costs a few dollars and a week. Listing an untested product costs your store's reputation when quality is bad or sizing runs small. Every product should pass through your hands or your agent's camera first. Pre-shipment quality checks are the cheapest insurance in this business.

Ignoring the duty math until it hurts. Sellers who priced products before the de minimis changes and never repriced are donating margin to customs. Audit landed costs quarterly, and reprice or delist anything that no longer clears your margin target.

Selling trademarked or counterfeit goods. Platforms and payment processors shut stores down for IP violations with little warning, and reinstatement is rarely quick. If a product carries a brand name you don't own the rights to, don't list it.

Supplier disappearance. AliExpress sellers close shop, change specs, or raise prices without notice. Keep a backup supplier for every winner, and once volume justifies it, move to a relationship-based supplier or agent who answers messages and holds stock.

No plan for returns. Shipping a $15 item back to China costs more than the item, so most sellers refund or reship cheap items without requiring a return. Decide your policy in advance and write it clearly on the store before the first complaint arrives.

When to move from dropshipping to bulk inventory

Dropshipping is a testing and cash-flow tool, not a permanent fulfillment strategy. The signals that it's time to buy inventory are concrete.

One SKU does 30+ orders a day consistently. At that volume, bulk buying, often 20 to 40% below dropship cost on 1688 or direct from a factory, compounds fast. A product doing 1,000 units a month at $4 saved per unit is $48,000 a year.

Shipping speed is costing you conversions. If analytics show abandoned carts citing delivery time, local inventory fixes the objection. Stocking proven winners in a US or EU warehouse, through a 3PL or a platform like CJ, typically cuts delivery to 2 to 5 days.

You want the product to be yours. Custom packaging, a branded insert card, a small design tweak only for you: none of this works well through marketplace dropshipping. It works once you own the inventory and control the spec. That is where real brand equity starts.

Refunds are eating margin. If your refund rate on a winner sits above 5%, a pre-shipment inspection on bulk orders usually pays for itself immediately. Catching defects before 500 units ship beats apologizing to 25 customers afterward.

Most sellers run a hybrid: bulk inventory for the two or three proven winners, dropshipping for everything still being tested. Cash stays tied up only where demand is proven.

Frequently asked questions

Is dropshipping from China legal in 2026?

Yes. What changed is customs treatment: low-value parcels no longer enter the US duty-free, and the EU is adding per-item duties on cheap imports. As long as goods clear customs properly, with correct declarations and duty paid, the business is fully legal. The sellers in trouble are the ones misdeclaring values or categories to dodge duty, which is customs fraud.

How long does shipping from China take now?

Standard AliExpress shipping to the US runs 7 to 25 days. CJ Dropshipping averages around 11 to 12 days, faster for products held in US warehouses. Agent-consolidated shipments from 1688 typically land in 7 to 15 days. None of these are next-day delivery, which is why setting expectations honestly matters more than shaving two days off transit.

Do I need a business license to dropship?

In most jurisdictions you can start as a sole proprietor, but requirements stack up fast: a registered business for payment processor verification, tax collection where you have nexus, and proper documentation for wholesale or agent relationships. Check local rules before scaling, not after your first tax notice.

How much money do I need to start?

A few hundred dollars covers a store subscription, a domain, and initial ad testing. Realistically, budget for samples, several rounds of ad testing at $10 to $20 a day, and your first refunds. Most sellers who make it work spend in the low four figures before finding a winner. The "start with $50" pitch ignores the cost of learning.

Can I dropship branded products from China?

Only with authorization. Listing trademarked brands without permission risks takedowns, payment holds, and store closure. Build your store on unbranded or private-label products you control.

Should I use an agent or stick with AliExpress?

Stick with AliExpress while testing, because a failed test should cost almost nothing. Bring in an agent when you have a winner worth protecting: better 1688 pricing, inspection before shipment, custom packaging, and someone who answers when a shipment goes wrong. If that is where you are, email hi@cnally.com with your product and monthly volume.

The decision rule for starting in 2026

Dropshipping from China in 2026 rewards sellers who treat it as a real supply chain, not a loophole. What remains is a legitimate model with tighter math: test cheaply on AliExpress, scale winners on faster fulfillment like CJ, recover margin through 1688 agents once volume justifies it, and move proven products to bulk inventory with proper QC before refunds force the decision.

Before you list your first product, run this filter: landed cost under one-third of your planned sell price, a shipping time you can state honestly on the product page, a supplier you have actually ordered a sample from, and a returns policy written before the first complaint. If a product passes all four, test it. If it doesn't, move on to the next one.

When you are ready to move past marketplace fulfillment, a sourcing partner on the ground in China changes the equation. CN Ally finds factory-direct suppliers, negotiates pricing, runs quality inspections before goods ship, and coordinates freight so your winners arrive on time and as described. Email hi@cnally.com with your product and monthly volume for a straight answer on whether the numbers work.