Individual standardized test scores in major school subjects can be combined to evaluate school programs and systems. Recently, Sweden experimented with such a practice. Test scores in different subjects (Swedish, English, mathematics) were gathered, units of different size (schools, municipalities, regions) were compared, and inter-school variances, inter-municipality variances, and inter-regional variances were determined. Problems with the use of individual assessment as a foundation for system evaluation include the fact that standardized test use in Sweden is voluntary and teachers, fearing that student results might be used to evaluate them as well as their students, might not use the tests. And if their use is made compulsory it might steer teaching and learning in an undesirable direction. Nevertheless, this method does have its uses. Taking the evaluation process a step further, data from an international evaluation study by the Association for the Evaluation of Educational Achievement showed that the between-school variance was lower in Sweden than in some other comparable countries. "Good" and "bad" student results were as frequent in Sweden as in the other countries, but were not as concentrated in special "good" schools and "bad" schools. (KH)