A study was conducted to investigate the intra-seasonal climate variability and change in semi-arid eastern Kenya and also assessed the ability of the households to discern trends in climate and how the perceived trends converge with actual long term weather observations. The study utilised long term climatic data and data collected through interviews of 200 households using a structured questionnaire. The survey data was analysed through descriptive statistics using the Statistical Package for Social Sciences (SPSS) version 12.0. The results of long term climatic data indicated high year-to-year variation in seasonal rainfall with 49.0% and 58% negative anomalies observed in the long and short rainfall seasons respectively. No discernible increasing or decreasing trend in the long- seasonal rainfall was observed over the period of study. However, Long-term temperature data showed high year-to-year variation in annual mean maximum and minimum temperatures with maximum temperature increasing during the period. Long term rainfall data (51 years) showed that 31.4 and 35.3% of the long rains would be classified as good and failed seasons respectively, with the remaining percentage classified as moderate seasons. For the short rains, 15.7 and 43.1 % would be classified as good and failed seasons respectively, with the remaining percentage classified as moderate seasons. Farmers interviewed were able to recollect the past seasons fairly accurately especially the 'good' and 'failed' seasons which corroborated well with the meteorological records. Indigenous knowledge on weather forecasting was reported by 81% of farmers to be helpful in farming decision making especialy on the types of crops to be planted. A better understanding of farmers' perceptions of climate change, ongoing adaptation measures, and the decision-making process would important to inform policies aimed at promoting sustainable adaptation of the agricultural sector.
