Bali cows (Bos javanicus) are a common breed in Indonesia and are favoured by smallholders for their size and ease of husbandry. However reproduction rates can be low in some areas and poor nutritional and reproductive management including poor mating management, low availability of genetically superior bulls, poor oestrous detection, poor Artificial Insemination (AI) procedures and poor availability of inseminators at the proper time are crucial issues which are likely to affect this. This thesis examined factors affecting post-partum anoestrus (PPA) and inter-calving interval (ICI) in Bali cows and heifers. This was examined by using large data sets to identify the factors which affect pregnancy rates in villages, developing methods to detect oestrus and pregnancy in villages, examining the effect of body condition score (BCS) on hormonal patterns and by quantifying milk production in cows.A large scale analysis of records from across the eastern islands of Indonesia indicated that there was a median inter-calving interval of 380 days but that postpartum anoestrus could not be determined precisely because of poor observations of first return to oestrus and mating. These records used observational data over 2-4 years where cows under various village management procedures were observed for weight, BCS and at least dates of calving from which various reproductive parameters could be calculated. ICI proved to be most useful and accurate given the quality of observed data. The range in ICI was very large and low annual calving % were recorded. There was an association of high reproduction rate with body condition score (BCS) at calving such that a BCS >3.5 (out of 5) was associated with intercalving intervals of approximately 380 days. The median ICI across all sites was 380 days but some sites recorded 75 percentile values in excess of 500 days.In controlled pen experiments, faecal progestogens and vaginal electrical conductivity were examined as techniques to detect oestrus and pregnancy in the field. Faecal progestogens could be used successfully but there were problems with vaginal electrical conductivity in routine application and this requires more work. Cycling heifers were observed over complete cycles and plasma progesterone and faecal progestogens examined every 2 days. The pattern of faecal progestogen iii concentration over an oestrous cycle mirrored that of plasma progesterone and hence could be used to monitor the oestrous cycle and pregnancy but the sensitivity was low and a number of samples would need to be taken to detect the cycling pattern and/or pregnancy. Similarly cows were monitored over the oestrous cycle and vaginal electrical conductivity measurements made but the association was not strong or sensitive enough to be used routinely. Associations between plasma leptin or plasma IGF-1 concentration and BCS were examined in another controlled pen experiment and plasma IGF-1 concentration was increased in response to higher BCS but plasma leptin concentration was not affected. Heifers ...