This paper describes the development of a very simple micro-flow pneumatic ICPMS nebulizing solution based on a commercial glass concentric slurry nebulizer and a fused-silica capillary. It appears as a well-performing, low cost system able to nebulize efficiently at a low liquid flow rate (40000 cps for a 1 ppb 238U solution at 1 μL min−1). Experimentally, it provides a practical set-up for studying the influence of not only operating conditions (liquid and gas flow rates, temperature) but also nebulizer design (nozzle configuration, capillary tip shape, capillary inner diameter). More precisely it allows the study of liquid-capillary-tip to gas-exit stick-out, i.e. recessed position (e < 0), coplanar liquid and gas exits (e = 0) or liquid protruding configuration (e > 0) in combination with various liquid capillary inner diameters or tip shapes. This investigation brings out several practical results even if the relation between the primary aerosol characteristics (studied by three complementary techniques (PIV, shadowgraphy and ILIDS) and the ICPMS signal measured on a tertiary aerosol is not easy to establish precisely. These results are systematically compared to the literature data, replacing them in a more global perspective. It highlights the prime importance of the nozzle configuration i.e. stick-out and liquid capillary inner diameter. The stick-out (e) parameter is of crucial importance both for the primary spray homogeneity in terms of droplet velocity distribution and for tertiary signal intensity and stability but shows no influence on the oxide ratio or on the droplet size distribution. This last characteristic may not be the most relevant criterion for low liquid uptake nebulizers, contrary to millilitre flow-rate ones. The authors hypothesize that this may be linked to the evaporation capacity of the nebulising gas largely exceeding the liquid uptake, creating a fundamentally different situation for the two types of nebulizers. This study also displays the interest of using a flat-end capillary instead of a tapered end which is the solution most commonly employed, and it highlights the existence of an optimum nozzle configuration (e > 0) in terms of sensitivity and stability, that is dependent on the capillary inner diameter.