In this paper I assess the potential for improvement in the documentation of metal detecting in the Swedish contract archaeology. I then investigate how such an improvement could be achieved. The study consists of three parts. I begin by defining what makes up high-quality documentation of metal detecting. I then use this definition to develop criteria for an assessment system. In the second part I use this assessment system for a quantitative analysis of archaeological archive reports from four of the largest contract-archaeology organizations in the Lake Mälaren area. Finally I report on interviews I have made with representatives from two of the studied organizations to seek possible explanations for the results of my analysis. My study confirms that there is considerable room for improvement in the analyzed archive reports, and sheds light on what parts of the documentation show frequent flaws. The results suggest that information specific to metal detecting is the most lacking, such as the selection of metal indications and fieldwork conditions. But there are also considerable flaws in more elementary information about surveys, which would probably never be accepted if it were a question of more traditional fieldwork methods.