Modern varieties of agricultural crops, commonly referred to as high yielding varieties (HYVs), are being increasingly looked upon as a major source of the additional foodgrain required to attain self-sufficiency in a number of Third World countries. In spite of the demonstrated productivity gains with HYVs, the adoption rate has been rather disappointing in many of those countries. Part of the reason behind the farmer's reluctance must arise from the unfamiliarity and high costs of the new technology (increased cost of seeds, fertilizer, pesticides, and irrigation). One way of conceptualizing this would be to recognize that HYVs are riskier than traditional varieties. Even though the vagaries of nature may affect all crops similarly, the higher costs incurred in planting HYVs result in bigger losses in times of poor yields. Thus although the potential return is higher, the possible losses are also higher. Lack of familiarity with the new technology may, in itself, be a source of subjective risk (Lipton). In this paper we show that a state run crop insurance programme may play a major role in the farmer's ability to take risks and allocate more land to cultivation of HYVs. The reasons for modelling a state run scheme are several. At the theoretical level, the recent literature suggests that insurance simply may not be provided by competitive markets due to imperfect information about the risk classes to which different farmers belong. At an empirical level, private insurance companies do not exist for agriculture alone. Further, private insurance companies, where they exist, use a large part of their budget to sell insurance. Although in a competitive environment one would expect that the increased social benefits from competition would outweigh the transaction costs, in view of the continual fight against starvation in many developing countries, it seems needless to waste resources selling insurance. There are thus sound theoretical and practical reasons for modelling crop insurance as a national programme. The paper suggests that the main thrust of the theoretical findings has already found empirical support from whatever scarce evidence is available. In this light, prospects for a state run insurance programme in terms of food selfsufficiency are discussed for Bangladesh.