A mechanistic upscaling scheme was developed to estimate temporal nitric oxide [NO] emissions from soils for three distinct land use classes by spatial extrapolation of laboratory measurements for Zimbabwe (383,667 km2). Laboratory measurements on miombo woodland, grassland, and agricultural Zimbabwean soils were used to derive moisture- and temperature-dependent NO emission algorithms. By combining monthly fields of modeled soil moisture and temperature (based on climate means, vegetation indices, soil, and terrain elevation data) a mean monthly soil NO flux, including rainfall-induced elevated flux events (pulsing), was obtained. Countrywide emission rates ranged from 0.1--0.4 ng N m-2 s-1 for the dry season to wet season mean fluxes of 3.7--9.4 ng N m-2 s-1 for miombo, 4.4--7.0 ng N m-2 s-1 for grassland, and 4.6--10.9 ng N m-2 s-1 for agriculture. Annual net soil NO emissions (less canopy reduction) for Zimbabwe were 32.9 Gg N yr-1, with miombo woodlands (66% of Zimbabwe) contributing 63% to the regional NO budget. Rainfall-induced NO pulsing contributed an additional 14% to countrywide annual bulk NO emissions. The majority of these pulses occurred during the first half of the rainy season (November--December), with agricultural soils contributing the most and grassland soils contributing the least to pulsing. The upscaling approach provided a first countrywide estimate of spatiotemporal soil NO emissions for a hitherto largely unresolved source: the soil-moisture-limited African subtropical savannas. ¿ 2001 American Geophysical Union