If you find Ancestry's indexes 90% reliable, you are doing a LOT better than most! The error rate for my own one name study - on surnames alone - is close to 20% - and as for place errors and other mistranscriptions I've not bothered to count, but they are both absurd and practically innumerable - (such as "Romania" for Raw (NBL) for instance).. The census CD's I publish in my Drake Software guise continue to sell reasonably well: the indexes are much more comprehensive than Ancestry, wildcard searching is possibly on any field and on the upcoming OXF51 I anticipate allowing searches for up to 3 forenames of persons sharing the same household, surname unknown. The data can also be arranged to point to either a DVD with the images if available, or to online images. All my programs have comprehensive mapping down to parish level within a county, and sometimes to even finer granularity, and permit of a number of types of statistical analysis of populations etc. But most important of all, the transcriptions are the work of the local FHS's, have been rigourously checked and are fully annotated to describe both anomalies and peripheral notes by the enumerator. You dont get any of that on line! Not that the Ancestry /images/ are to be decried - their quality is in general good, and if you can manage to penetrate their abysmal indexing, most useful - i have now donwloaded around 2000 in pursuit of my AINSLEYs. So - the pure image CD without an index is probably struggling - but GOOD transcriptions with added value in the form of comprehensive search engines and analysis have a place for the serious genealogist and local historian - certainly enough to encourage me to keep on producing them! Hugh Ainsley - Drake Software and the AINSLEY one name study - GOONS #3926