第一类备案与第二类注册的检测要求差在哪

结论：差别不只是项目多少，证据的形式要求也不同

一类备案与二类注册在检测上的差异，常被简化成「一类测得少、二类测得多」。项目数量确实有差别，但更关键的差异在于证据的形式要求：谁出的报告算数、报告需要满足什么条件、数据要覆盖到什么程度。

这个差异决定了准备工作的性质完全不同。按一类的思路准备二类的资料，即使项目都测了，形式上也可能不被接受。

两者的主要差别

第一行是根本差异。一类产品通常可以凭企业自身的检验能力出具报告，这意味着企业要有对应的设备、方法和人员，并且检验过程要可追溯。「自检」不等于「随便测」——自检报告同样需要方法依据、设备溯源和原始记录，只是不要求第三方出具。

实际中比较常见的问题是，企业以为一类产品可以简单处理，检验记录做得很粗，等到监管检查时拿不出支撑数据。

项目范围的差别在哪

一类产品的检测通常围绕产品技术要求里列的性能指标展开。这些指标由企业自己制定，所以范围在一定程度上可控。

二类产品除了性能指标，还要覆盖适用的安全标准和产品专用标准。这部分不由企业选择，而是由产品属性决定。判断哪些标准适用，是二类产品检测准备里最容易出错的一步。 漏掉一项适用标准，往往要到技术审评阶段才被指出，那时候补测的时间成本很高。

分类判断错误的代价

分类判断错误有两个方向，代价不对称。

按一类准备、实际应属二类。 这个方向代价很高。备案拿到之后被认定分类错误，产品可能面临停止销售，重新按二类走注册流程，而检测要从头做——因为原来的自检报告在形式上不满足二类要求。

按二类准备、实际属一类。 这个方向浪费了一些钱和时间，但产品本身没有风险，资料只多不少。

所以在分类不确定时，按较严的口径准备是更稳的选择。多花的检测费用，相比分类判断错误的后果，通常是划算的。

处于分类边界时怎么办

有些产品的分类确实存在判断空间，比如功能上介于两类之间、或者新出现的产品形态尚无明确归类。

处理思路有三条：

一是查已上市同类产品的分类。 找功能相近、预期用途相近的已上市产品，看它们按哪一类管理。这不是决定性依据，但有参考价值。

二是走分类界定程序。 通过正式渠道申请界定，拿到明确答复。周期较长，但结果确定。

三是按较严口径准备，同时申请界定。 两条线并行，界定结果出来时检测已经做得差不多了。这样时间上不吃亏，代价是如果界定为一类，多做了一些项目。

检测这一环建议按第三条走。 因为检测周期长，等界定结果出来再开始，时间上往往来不及。

技术要求的写法差别

一类产品的技术要求由企业自定，但仍要做到指标可测、判定清晰。二类产品的技术要求除了自定指标，还要纳入适用标准的强制性条款，并且指标值不能低于标准要求。

实务上容易出的问题是：企业把标准里的条款照抄进技术要求，但抄的是不适用于自己产品的那部分，结果检测时按技术要求测，测出一堆本来不必测的项目，还可能因为不适用而判不合格。技术要求里写的每一条，都应当是真正适用且打算被检测的。

自检能力的实际门槛

一类产品可以自检，但「能自检」是有门槛的。至少要具备：对应的检测设备且在有效校准状态；经过确认的检验方法；经过培训授权的检验人员；完整的原始记录与报告管理制度。

这几项里，设备校准和原始记录是检查时被看得最多的。设备有但校准过期、测了但没有原始记录，等同于没有检验。 有些企业买了设备就认为具备了自检能力，忽略了后面这套管理要求。

如果评估下来自身条件不具备，委托第三方检测同样可行，一类产品并不禁止委托检验。这时候成本会上去一些，但形式上更稳妥。

产品升级带来的类别变化

还有一种情形值得提醒：产品在迭代过程中增加了功能，可能从一类变成二类。

典型的例子是给原本纯机械的产品加上电动驱动或者数据采集功能。企业往往按「只是加了个功能」来理解，仍按一类处理，而实际上产品属性已经变了。

判断的触发条件是：新增功能有没有引入新的风险类型。 引入了，就要重新评估管理类别，而不是沿用原来的判断。

我们的做法

我们接到委托时会先问产品的管理类别，以及这个判断的依据是什么。如果委托方自己也不确定，我们会建议按较严的口径先排项目清单，并把「如果是一类可以省掉哪些」单独标出来。

这样做的好处是，委托方拿到的是一份带判断依据的清单，而不是一个数字。分类的最终判断不在实验室这一端，但检测方案可以做成对两种结果都有准备的形式，不至于因为分类结论变化而推倒重来。

如果你的产品分类还没定，或者不确定哪些标准适用，可以把产品资料和预期用途描述发过来一起看，或者直接联系：132 4819 8029。能力范围见服务介绍，送检要求见送检要求，更多内容见知识库。

English version

Conclusion: the difference is not only how many items, but what form the evidence must take

The difference in testing between Class I filing and Class II registration is often reduced to Class I requiring fewer items. The number of items does differ, but the more consequential difference is in the form the evidence must take: who may issue the report, what conditions it must meet, and how far the data must reach.

That difference makes the preparation work fundamentally different in kind. Preparing a Class II dossier with a Class I mindset can fail on form even when every item has been tested.

The main differences

For report issuance, Class I generally permits testing by the manufacturer itself, while Class II requires a body with appropriate accreditation. For scope, Class I covers the performance indicators in the product technical requirements, while Class II adds applicable safety and product-specific standards. Coverage justification is simplified for Class I and must be argued item by item for Class II. Biological evaluation is comparatively basic for Class I and systematic for Class II according to contact nature and duration. Review is largely formal for Class I and substantive for Class II. Change handling is simpler for Class I.

The first of these is the fundamental one. Class I products can generally rely on the manufacturer's own testing capability, which means having the equipment, methods and personnel, with a traceable process. Self-testing is not the same as casual testing: a self-issued report still needs a method basis, equipment traceability and original records; it simply does not require a third party.

A common problem is manufacturers assuming Class I can be handled loosely, keeping thin records, and then being unable to produce supporting data during an inspection.

Where the scope differs

Class I testing generally addresses the performance indicators the manufacturer has set in its own technical requirements, so the scope is to some extent within its control.

Class II additionally covers applicable safety standards and product-specific standards. That part is determined by the nature of the product rather than chosen. Determining which standards apply is the step where Class II preparation most often goes wrong. A missed applicable standard typically surfaces during technical review, when the cost of catching up is high.

The cost of misclassification

The two directions of error are not symmetric.

Preparing as Class I when the product is actually Class II is expensive. If the classification is found to be wrong after filing, the product may have to stop being sold and go through registration from the beginning, with testing restarting because the original self-issued report does not meet Class II requirements in form.

Preparing as Class II when the product is actually Class I wastes some money and time, but carries no product risk and leaves the dossier over-supported rather than under-supported.

Where classification is uncertain, preparing to the stricter reading is the safer choice. The additional testing cost is usually small against the consequences of getting it wrong.

Products on the boundary

Some products genuinely sit in a grey area, whether functionally between two classes or as a new form with no settled classification.

Three approaches help. Look at how comparable marketed products are classified; not decisive, but indicative. Apply through the formal classification determination route for a definitive answer, accepting a longer timeline. Or run both in parallel: prepare to the stricter reading while the determination is pending.

For testing specifically, the third approach is usually right, because testing lead times are long and waiting for a determination before starting often leaves too little time.

Writing technical requirements

Class I technical requirements are set by the manufacturer but still need measurable indicators and clear acceptance criteria. Class II requirements must additionally incorporate the mandatory clauses of applicable standards, with values no less demanding than the standard.

A frequent error is copying clauses from a standard into the technical requirements including parts that do not apply to the product. Testing then follows the technical requirements, producing items that need not have been tested and that may be judged nonconforming precisely because they do not apply. Every line in the technical requirements should be genuinely applicable and intended to be tested.

How we handle it

We ask what class the product falls into and what that judgement rests on. Where the client is unsure, we suggest scoping the item list to the stricter reading and separately marking what could be dropped if it turns out to be Class I.

The client then has a list with reasoning attached rather than a single number. The final classification decision does not sit with the laboratory, but a test plan can be structured to accommodate either outcome without starting over.

If your product's classification is unsettled, or you are unsure which standards apply, send us the product information and intended use statement. Phone or WeChat: +86 132 4819 8029.