currently, ValidationTagLib's hasErrors closure does not support the model attribute very well. For instance, the following code: <g:hasErrors model="[a:a, b:b]"> <g:eachError bean="${<here is the problem>}">${it.defaultMessage}</g:eachError> </g:hasErrors> The problem with this code is that the body of <g:hasErrors> will be iterated over twice (because we have 2 objects in our model). The inner body does not have access to the current model object that's being iterated over. The relevant code that's in question is here: def hasErrors = { attrs, body -> //................................. if(errors) { if(attrs['field']) { if(errors.hasFieldErrors(attrs['field'])) { out << body() } } else { out << body() } } } } The code does not pass the current error object being iterated over. Hence, I am unable to access and print out the right error object codes. The fix is to pass the current error object being iterated over to the body for use, as follows: def hasErrors = { attrs, body -> //................................. if(errors) { if(attrs['field']) { if(errors.hasFieldErrors(attrs['field'])) { out << body(errors) } } else { out << body(errors) } } } } This will allow the body to do something like this: <g:hasErrors model="[a:a, b:b]"> <g:eachError bean="${it}">${it.defaultMessage}</g:eachError> </g:hasErrors> Now the eachError can properly access the current error object in the model and display each error.