Climate change and variability pose serious risks to rain-fed agricultural land use in the semi-dry agro-ecological zones of Nigeria. Kogi State, is not an exception. Rainfall is becoming more unpredictable and unreliable both in its timing and its volume and growing seasons are changing, ecological zones are shifting. Therefore, this study was conducted to find out the impact of climate change on agricultural land use in Kogi State in order to implement better adaptation strategies for smallholders. For this purpose, forty years (1975-2014) rainfall and temperature data from NIMET, Landsat images (1987, 2001 and 2014) and questionnaire on farmer’s perception were used. Focused variables calculated were: Standardized anomalies, means for both temperature and rainfall. Landsat images were used to assess Land Use/Land Cover (LULC) types between 1987 and 2014. A questionnaire was administrated to investigate farmer’s perception on climate change and agricultural land use. Data collected were analysed using SPSS 20. Excel 2013 and Envi5.1. The magnitude of the change rate between 1987 and 2001, 2001 and from 2987 to 2014. It obvious that only water body did not experience a major change. It was observed that a decrease in natural vegetation within the period of 1987-2001 (15) and from 1987 to 2014 (29). This indicate the practice of deforestation in the study area for residential purpose and farming. Binary Logit Model was used to determine the significant factor that affects farmer’s decision to change land use under cultivation. Results show a slight increase in both temperature and rainfall up to 0.4 and 0.10 respectively. Lowest and highest temperatures were recorded in 1989 and 2005 respectively with values of 27.30C and 29.20C. The period of 1975 to 1997 is characterized as the cool years as the values were below the normal (zero) except years 1980, 1983, 1987, 1990 and 1996, while from 1998 to 2013 mean annual temperature was above the normal. The average rainfall is 1218.5 mm. The study area has experienced successive dryness years from 1975 to 1985 with a wet year occurring in 1978. The dry year was 1977 while the wetness was 1999 with respective standardized value of -1.82 (moderate dryness) and +2.23 (extreme dryness). The statistic of land use and land cover map shows that between 1987 and 2001 (15 years), the built up area had increased by 41.565 hectares. While farmland decreased by -27577 hectares between 1987 and 2001. Between 2001 and 2014 only built up area increased tremendously by 10.68% within the period. In terms of adaptation strategies, seven (7) adaptation strategies were employed by farmers, viz: changing planting dates, implement soil conservation schemes and change crop variety at 31%, 22% and 21% respectively in response to change in rainfall. Logit regression model reveals that change in rainfall patterns (Fpr. = 0.122) has no significant impact on farmers’ decision to change land use under cultivation at 5% level of significance. However, number of family member schooling (Fpr. = 0.01), household farming experience (Fpr. = 0.00) and access to credit (Fpr. = 0.00) have significant impact on farmers’ decision to change land under cultivation. These results will be useful for agricultural planning in the study.