Returning 404 is the intent

This particular API was created when "RESTful" interfaces were in vogue. Although there is no formal spec for designing these types of APIs, my understanding is that it's generally considered a good idea to make full use of the available range of HTTP status codes to accurately relay information about the nature of a request.

When requesting a resource (like in this case, a postcode) it's common to return a 404 response code if that resource does not exist. A good indicator that you're dealing with a resource is if the URI looks something like /resource/:resource_id

I don't think postcodes.io is in the minority here, most HTTP API's designed within the last ~5 years take a similar approach. The Ruby on Rails framework, which is built around RESTful ideas, returns 404 by default for a non existent resource. Likewise (afaik) many popular REST API's like Twitter and Stripe exhibit similar behaviour.

However, if you were performing a query like /postcodes?q=s625ty I'd agree with you that a 200 status code should be returned for both a populated and empty result set.

It's only /outcodes/:id and /postcodes/:id that exhibit this behaviour as they are considered as resources. All other "query-like" methods will return the 200 status code you're expecting.

Sounds right.