Abstract An experimental and inverse computational study is presented of spray cooling of microelectronics with an emphasis on the spray angle effects on cooling performance. A thermal test chip provides the heated target, and is cooled by a single pressure swirl atomizer. Thermal readings were taken at the spray angles of 0–60°, at a fixed distance of 1.4 cm from the heated die surface. An inverse heat transfer computational algorithm is developed to calculate the unknown spray cooling heat fluxes using the measured temperature data inside the die. The computational scheme is a combination of the finite element method and the truncated single value decomposition with the discrepancy principle for determining the optimal truncation threshold value. Good agreement is obtained between the experimental measurements and calculated results. For this particular system, a direct estimate using temperature readings at two adjacent points would produce incorrect heat flux results and an inverse algorithm is deemed essential if an accurate heat flux is to be obtained from the measurements. It is found that a major cause for the drop-off is the reduction in spray volumetric flux delivered to the die at the greater spray angles.