BACKGROUND Pragmatic assessment methods are very diverse and differ in informant type. Some rely on parents, others on teachers/professionals and some directly test pragmatic abilities in the children themselves. A widely used pragmatic parent questionnaire is the Children's Communication Checklist--2 (CCC-2). However, it is not known how scores on the CCC-2 relate to direct measures of pragmatics. OBJECTIVE The aim of the current study is determine whether children's language patterns on pragmatics obtained with a parent questionnaire were converging with findings when the children were directly tested with a pragmatic test. METHODS The CCC-2 and the Nijmegen Pragmatics Test (NPT) were applied to 24 pre-schoolers (aged 4-7 years) with various language impairments and 33 age-matched typically developing pre-schoolers. RESULTS Both pragmatic language instruments clearly differentiated between pre-schoolers with language impairments and those without language impairments. However, the obtained correlations between the different measures were low to moderate. The specificity of each of the instruments was sufficient, but the sensitivity was generally poor. CONCLUSIONS The instruments were not always converging, but when the instruments did converge the obtained results were valid. However, the obtained high specificity and relatively low sensitivity values for each of the instruments showed that better cut-off scores are needed. When only one of the instruments indicated the absence or presence of language impairments, one needs to be careful in concluding whether or not there are indeed language impairments.