The highly variable rainfall and recurring drought in dry sub-humid and semi-arid regions constitute a major socio-economic problem that influences people's livelihoods. Agricultural drought is the most common of all drought types and is the most challenging to track and forecast owing to its ease of occurrence and impact on socio-economic sustainability. In Nigeria as is the case in large parts of sub-Saharan Africa, drought frequency, intensity and effect have dominated the literature since the great Sahelian drought of the late 1960s and early 1970s. This study proposes a simple drought monitoring and early warning (EW) methodology based on an experiment over Kano, Nigeria, hinged on an Intraseasonal Rainfall Monitoring Index (IRMI, Usman and Abdulkadir, 2012) determined for each pentad, using daily rainfall totals for a period of 30 years. The onset dates and drought intensity levels were used to define EW statements at three levels: advisory, alert and emergency. An experimental application of the scheme identified and detected late initiation conditions of real monsoon onset (RMO) in those seasons that eventually witnessed disastrous rainfall situations. Generally, the result reveals that the years between 1970 and 1990 over Kano, Nigeria, which ended as years of severe drought or worse, showed enough signals of poor rainfall distribution (deficient moisture) at the beginning for the scheme to have enabled a very high degree of forecast accuracy. It is concluded that strict monitoring of the rainfall regime especially during its onset phase is capable of revealing danger signals in the rainfall regime early enough to enable mitigation. Such mitigation is critical for the management of agricultural productivity in the predominantly rainfall-dependent farming systems in the Sudan-Sahel zone as confirmed by earlier studies.