In vineyards, at early growth stage, pesticide spraying is horizontal and is usually performed without air flow assistance. However, nozzles are characterized by a global value, namely the Volume Mean Diameter, measured for a vertical spray. In this study, we compared the spatial distribution of droplet size and velocity issued from different nozzles when spraying vertically and horizontally, in order to assess which nozzles are most appropriate. Three different nozzles delivering the same flow rate have been chosen, namely an air-induction flat fan nozzle, an air-induction hollow cone nozzle and a classical hollow cone nozzle, the last one corresponding to the conventional nozzle type used in vineyards. Phase Doppler Particle Analyzer measurements have been performed at 40 cm from the nozzle exit within a horizontal plane and a vertical plane, with the nozzles spraying vertically and horizontally respectively. Each plane consisted of a hundred of measurement points. The axial velocity component and the droplet diameter measurements were performed simultaneously in each measurement point. In these experiments, the air-induction flat fan nozzle delivered fewer droplets than other nozzles, with a larger size and a higher velocity, despite a low injection pressure. Results have also shown that the proportion of droplets with a diameter lower than 100 microns corresponded to 14 % for the air-induction flat fan nozzle, 30 % for the air induction hollow cone nozzle and 68 % for the classical hollow cone nozzle. Moreover, the part of droplets with an axial velocity lower than 2 m/s was equal to 80 % for both hollow cone nozzles whereas it was equal to 40 % for the air induction flat fan nozzle. This velocity criterion was in accordance with the conclusions of another experimental study which showed that maximum deposition on an artificial crop was obtained for the flat fan air-injection nozzle, when compared with the other two nozzles (Naud et al, 2014). Lastly, results have shown that the Volume Mean Diameter was very slightly modified by the nozzle orientation. However, the spatial representation of the mean diameter within the measurement plane differed greatly when the nozzle sprayed horizontally compared to the vertical spraying. Gravity effect was noticeable. To conclude, it was demonstrated that when spraying horizontally (which is the case in vineyards), nozzle classification should integer axial velocity aspect because too slow droplets may contribute to off-target deposition. The air-induction flat fan nozzle corresponded to the most appropriate nozzle in vineyards at early growth stage. In a near future, Phase Doppler Particle Analyzer measurements will be carried out using an air flow added to the liquid spraying, in order to assess other growth stages.