There are two possible explanation: (1) your neutralizing assay did not work; or (2) your antibodies do not recognize neutralizing epitopes. Did you include a positive control for your neutralizing assay, ie. an antibody that is known to neutralize infection? Antibodies that work on one assay do not have to work on another assay because they recognize different epitopes. This is why you have to test your antibodies on all assays. For example antibodies that work on western would recognize a linear epitope whereas antibodies that neutralize infection have to recognize neutralizing epitopes.

I think Pei-yin is right and I'd be tempted to go with his second reason.  Polyclonal antibodies are a mixture of so many antibodies that it's very difficult to figure out what they contain. you must have used an in-vitro expressed  protein to raise the antibodies. Sometimes if your protein is not purified well- it may contain some cellular proteins against which a lot of your polyclonal antibodies will be. Sometimes if the majority of the neutralising epitopes in your protein are conformational and you've an in-vitro expressed protein- it'd not have the required secondary/tertiary structure to expose the epitopes to the antibody producing cells- this also happens if your protein is somehow degraded; and you may end up with lots of non-neutralising antibodies.

Thanks for the answer that kindly provided by you all.

I have included the positive control which is the serum and it showed that it successfully neutralize the virus to certain extent, so the neutralization assay shouldnt have problem.

 Which virus are you testing?

coxsackievirus A16

This one is relatively stable one. But neutralization assays is artificial model which might work for different epitops in vitro and in vivo