The area chosen for this study is the Central Elburz of Iran surrounding the southern end of the Caspian Sea. It includes an important rainfall dependent agricultural area and merges into the semi arid zone with a well defined boundary. From social and agricultural points of view rainfall is by far the most important climatic factor in many tropical and subtropical countries. An important aspect in the development of agriculture in the Central Elburz is the determination of the seasonal rainfall patterns. This thesis should be considered as a contribution to the study of the seasonal rainfall regime of Iran, with direct value to agriculture in the region. The purpose of this thesis is to examine the pattern of rainfall especially in relation to the growing season both in time and spatially. Also the thesis describes the variation in the availability of moisture for growth throughout this region. The thesis examines annual, monthly and seasonal rainfall over the region. Multivariate analysis has demonstrated that the study area can be divided into three rainfall regions. Before considering any analysis which could allow prediction on probability of future rainfall amounts, it is important to consider whether or not there on trends or fluctuations. The result of this analysis is that, few stations show positive trends, others negative trends. Rainfall variability has been quantified by the coefficient of variation index and analysed in a similar way to that of rainfall distribution. The results for most stations shown a relationship between variability indices and rainfall amounts. In relation to the study of variability, an assessment of rainfall probability and reliability has been considered with reference to rainfall critical for crops. Probability has a particular value in its application to agriculture. In this study 90% probability is examined and this is related to the economy of the agriculture units in the Central Elburz. Also 80 and 75 percent probability are presented. These spatial patterns of probability thus provide maps of agricultural potential. Rainfall records are analysed to provide estimates, percentage points of rainfall totals, variation in the start, and end of the rainy season and or the probability of dry spells within the rainy season. The estimation of potential evapotranspiration which are discussed and explained in this study are based on the Blaney- Criddle and Pan evaporation methods. The water balance approach provides the most rational method of analysing seasonal pattern since it considers characteristics of the dominant crops in this area, wheat and citrus, and it examines effectiveness of rainfall for these crops. Blaney- Criddle and Pan evaporation have been used to estimate potential evaporation for wheat and citrus and compared to actual evapotranspiration to give potential water deficit. The adoption of probability analysis of potential water deficit provides a valuable description of water availability for crops in this region of variable rainfall.