Abstract
This paper treats quality control in a wide sense which is imperative for a statistical agencyto do. First, we discuss the meaning of total survey error and different quality characteris-
tics. Secondly, some common quality problems in statistical agencies are presented as a prelude to the bulk of this paper, namely the case of Sweden. Quality control efforts at Statistics Sweden are reviewed regarding the handling of specific error sources, error estimation, and error control. Different tools such as preventive control, production control, and evaluation are dealt with. As for QC-procedures proper we provide some applications from clerical coding.