The project combined PPB with marker-assisted selection (MAS) by continuing the marker-assisted backcross programme begun in project R6673 to introduce QTLs for root growth and drought resistance into Kalinga III. Bulks generated through marker-assisted backcrossing were used for participatory plant breeding. Farmers made selections in the bulks in 3 seasons and pure bulks were obtained which out-performed Kalinga III. A control bulk with no QTLs was just as acceptable to farmers as bulks with root QTLs. This indicates that other Azucena alleles contribute favourably in the backcross material. Advanced backcross lines containing root QTLs were screened for root traits in soil-filled pipes under field conditions and in growth-room conditions in order to test for the effects of QTLs. Results of these studies were inconclusive although the material studied was not necessarily fixed at all alleles which could have influenced the very large standard errors obtained. Crosses were made to pyramid root QTLs into single genotypes and seeds are now available for farmers which contain all four root QTLs in the Kalinga III genetic background. PPB in this project, and in other DFID-PSRP projects, was done using progeny from a number of different crosses and also with products from mutation breeding. Farmers in India and Nepal across a range of rice environments identified many successful lines or bulks. The project combined PPB with molecular markers by evaluating the products of PPB for marker variation. Initially AFLPs were used, but these were not useful for comparative mapping. Methods for increasing the through-put of microsatellites (simple sequence repeats, SSRs) were tested and this method was used to analyse successful PPB lines. Towards the end of the project the method of SNPs (single nucleotide polymorphisms) became available and a small number were used to test PPB lines and bulks. These studies have been used to illustrate graphically the parental contribution of certain individuals or sets of selected material. Marker analysis has started to reveal some of the effects of farmer selection at certain chromosome regions or individual markers. This has led us to develop the hypothesis that markers can be used to identify farmer-preferred chromosome regions. These could be cross-specific or ecosystem-specific. The method used for analysis of PPB products has been called marker-evaluated selection (MES). A strategy has been identified which will use MES to isolate makers which could be linked to traits of specific importance to farmers in certain ecosystems. They can then be used in MAS to improve existing lines which do not contain these markers so that they should perform even better in the target environment.