If microcomputers are used as automated testing stations, for use in psychometric assessment, there are economic advantages. Discussion follows, however, on whether it is possible to improve the quality of cognitive assessment by extending the range of cognitive abilities to be assessed. Two types of extension are considered: modifying and expanding testing procedures for psychological functions that are components of conventional tests, and the extension of testing to psychological functions not generally assessed by conventional intelligence or aptitude tests. Computerized presentations will make relatively little difference in the current ways of testing verbal comprehension. Computer controlled testing could well extend the ways in which spatial-visual reasoning and memory are evaluated. The impact of testing on the evaluation of reasoning is unclear. Computer-controlled item presentation makes it possible to conceive of tests of learning and attention. The psychological nature of the abilities being assessed raises problems in assessment that are not addressed by the fact of computer-control. Some research questions are identified that ought to be explored before testing is extended into these fields. Computer-controlled evaluation could be extended to the assessment of criterion performance, either in the normal working situation or in a simulation of the workplace. While evaluation of this sort does raise social questions, it clearly presents an opportunity to obtain validation data for psychological assessment studies. (Author/BW)